Support power supply technical index evaluation method and system for novel electric power system

Abstract

The invention discloses a supporting power supply technical index evaluation method and system for a novel electric power system, and the method comprises the steps: constructing the technical index of each supporting power supply, comprehensively covering the performances of each dimension of the supporting power supply, determining the subjective weight of each index through employing a dynamic network analytic hierarchy process, and carrying out the evaluation of the technical indexes. The objective weight of each index is calculated by using a weight fusion method based on an entropy method, a machine learning algorithm and grey correlation analysis, the actual contributions of the indexes in different scenes are comprehensively reflected, the subjective weight and the objective weight of each index are dynamically fused to obtain the comprehensive weight of each index, and the comprehensive weight of each index is calculated. According to the method, the influence of subjective weight and objective weight is effectively balanced, the improved TOPSIS method is used for sorting the plurality of supporting power supplies based on the comprehensive weight of each index of the supporting power supplies, and the technical index evaluation results of the plurality of supporting power supplies are obtained, so that the comprehensiveness, flexibility and simplicity of the technical index evaluation of the supporting power supplies are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of index evaluation, and particularly to a method and system for evaluating technical indicators of a supporting power source for a new power system. Background Art

[0002] Under the background of accelerating global energy transformation, the goals of carbon peak and carbon neutrality have promoted the rapid transformation of the energy and power system towards a new power system with new energy as the main body. The new power system takes the high-proportion access and consumption of clean energy as the core, and requires higher safety, flexibility and efficiency to meet the needs of high-quality social and economic development. However, the construction and operation of the new power system face many challenges, including the intermittency and volatility brought by the high-proportion access of renewable energy, the deep coupling of various energy forms, and the complexity of the electricity market mechanism and policy constraints. These challenges pose higher requirements for the planning, optimization and dispatching of the power system.

[0003] As an important part of the new power system, the supporting power source plays a key role in this context. It not only needs to ensure the frequency and voltage stability of the power grid when high-proportion clean energy is accessed, but also needs to provide peak shaving and frequency modulation capabilities to match the flexibility requirements of the new energy fluctuations for the system. At the same time, it promotes the green and low-carbon transformation by improving energy efficiency, optimizing energy configuration and reducing carbon emissions. However, the current technical evaluation system for supporting power sources is not yet perfect. Existing methods often only focus on single-dimensional performance indicators, such as economy or carbon emission reduction ability, lacking comprehensive evaluation of multi-dimensional characteristics. At the same time, some evaluation systems lack a scientific basis in index selection, are disconnected from the actual application scenarios, and are difficult to comprehensively reflect the comprehensive performance of the supporting power source. In addition, with technological progress and policy changes, existing methods are difficult to dynamically adjust to meet the needs of different scenarios, and the complexity of some methods limits their operability in actual projects, resulting in high data collection and calculation costs. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to provide a method and system for evaluating technical indicators of a supporting power source for a new power system, which improves the comprehensiveness, flexibility and simplicity of the evaluation of technical indicators of the supporting power source.

[0005] To solve the above technical problem, a technical solution adopted by the present invention is: A method for evaluating technical indicators of a supporting power source for a new power system, comprising the steps of: Constructing the technical indicators of each supporting power source, where the technical indicators include a clean and low-carbon primary indicator, a safe and abundant primary indicator, an economic and efficient primary indicator, a supply-demand coordination primary indicator, and a flexible and intelligent primary indicator; Determine the subjective weight of each indicator using the dynamic network analytic hierarchy process, and calculate the objective weight of each indicator using a weight fusion method based on the entropy value method, machine learning algorithms, and grey relational analysis; Dynamically fuse the subjective weight of each indicator with the objective weight of each indicator to obtain the comprehensive weight of each indicator; Use the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each indicator of the support power supply, and obtain the technical indicator evaluation results of multiple support power supplies.

[0006] To solve the above technical problems, another technical solution adopted by the present invention is: A technical indicator evaluation system for support power supplies for a new power system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Construct the technical indicators of each support power supply, where the technical indicators include primary indicators of clean and low-carbon, primary indicators of safety and adequacy, primary indicators of economic efficiency, primary indicators of supply-demand coordination, and primary indicators of flexibility and intelligence; Determine the subjective weight of each indicator using the dynamic network analytic hierarchy process, and calculate the objective weight of each indicator using a weight fusion method based on the entropy value method, machine learning algorithms, and grey relational analysis; Dynamically fuse the subjective weight of each indicator with the objective weight of each indicator to obtain the comprehensive weight of each indicator; Use the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each indicator of the support power supply, and obtain the technical indicator evaluation results of multiple support power supplies.

[0007] The beneficial effects of the present invention are as follows: The technical indicators of each supporting power source are constructed. The technical indicators include the first-level indicators of clean and low-carbon, safe and sufficient, economic and efficient, supply-demand coordination, and flexible and intelligent. The comprehensive evaluation considering multi-dimensional characteristics is carried out, covering all-dimensional performances of the supporting power source. And the subjective weight of each indicator is determined by using the dynamic analytic hierarchy process, effectively solving the problems of strong subjectivity and insufficient fuzziness in the traditional analytic hierarchy process. The objective weight of each indicator is calculated by using a weight fusion method based on the entropy method, machine learning algorithms, and grey relational analysis, which can comprehensively reflect the actual contribution of the indicator in different scenarios and adapt to the needs of policy changes and technological development. The subjective weight and objective weight of each indicator are dynamically fused to obtain the comprehensive weight of each indicator, effectively balancing the influence of subjective weight and objective weight, improving the adaptability and scientificity of the comprehensive weight calculation. Finally, the improved TOPSIS method is used to rank multiple supporting power sources based on the comprehensive weight of each indicator of the supporting power source, obtaining the technical indicator evaluation results of multiple supporting power sources, enhancing the dynamic adaptability of the evaluation model to complex multi-dimensional scenarios, reducing data processing and calculation complexity, and thus improving the comprehensiveness, flexibility, and simplicity of the technical indicator evaluation of the supporting power source. Description of the Drawings

[0008] Figure 1 It is a step flowchart of a method for evaluating technical indicators of a supporting power source for a new power system according to an embodiment of the present invention; Figure 2 It is a schematic structural diagram of a system for evaluating technical indicators of a supporting power source for a new power system according to an embodiment of the present invention; Figure 3 It is a schematic diagram of technical indicators of a supporting power source in a method for evaluating technical indicators of a supporting power source for a new power system according to an embodiment of the present invention; Figure 4 It is an evaluation flowchart in a method for evaluating technical indicators of a supporting power source for a new power system according to an embodiment of the present invention. Detailed Embodiment

[0009] To describe the technical content, achieved objectives, and effects of the present invention in detail, the following is described in conjunction with the embodiments and with reference to the drawings.

[0010] Please refer to Figure 1 , a method for evaluating technical indicators of a supporting power source for a new power system, including the steps: Construct the technical indicators of each supporting power source, where the technical indicators include the first-level indicators of clean and low-carbon, safe and sufficient, economic and efficient, supply-demand coordination, and flexible and intelligent; Determine the subjective weight of each index using the dynamic network analytic hierarchy process, and calculate the objective weight of each index using a weight fusion method based on the entropy value method, machine learning algorithms, and grey relational analysis; Dynamically fuse the subjective weight of each index with the objective weight of each index to obtain the comprehensive weight of each index; Use the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each index of the support power supply, and obtain the technical index evaluation results of multiple support power supplies.

[0011] As can be seen from the above description, the beneficial effects of the present invention are as follows: Construct the technical indicators of each support power supply, where the technical indicators include first-level indicators of clean and low-carbon, safe and sufficient, economic and efficient, supply-demand coordination, and flexible and intelligent. It considers the comprehensive evaluation of multi-dimensional characteristics, comprehensively covers the performance of each dimension of the support power supply, and uses the dynamic network analytic hierarchy process to determine the subjective weight of each index, effectively solving the problems of strong subjectivity and insufficient fuzziness in the traditional analytic hierarchy process. Use a weight fusion method based on the entropy value method, machine learning algorithms, and grey relational analysis to calculate the objective weight of each index, which can comprehensively reflect the actual contribution of the index in different scenarios and adapt to the needs of policy changes and technological development. Dynamically fuse the subjective weight and objective weight of each index to obtain the comprehensive weight of each index, effectively balancing the influence of subjective weight and objective weight, improving the adaptability and scientificity of the comprehensive weight calculation. Finally, use the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each index of the support power supply, obtain the technical index evaluation results of multiple support power supplies, enhance the dynamic adaptation ability of the evaluation model to complex multi-dimensional scenarios, reduce the data processing and calculation complexity, and thus improve the comprehensiveness, flexibility, and simplicity of the technical index evaluation of the support power supply.

[0012] Further, the determining the subjective weight of each index using the dynamic network analytic hierarchy process includes: Obtain the lowest possible value, medium possible value, and highest possible value of the relative importance of each index to another index determined by experts; Determine a triangular fuzzy judgment matrix according to the lowest possible value, medium possible value, and highest possible value of the relative importance of each index to another index; Improve the triangular fuzzy judgment matrix using a dynamic adjustment factor to obtain a dynamic fuzzy judgment matrix; Determine the membership function; Calculate the consistency index of the dynamic fuzzy judgment matrix, and calculate the random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index; Judge whether the random consistency ratio is less than a preset value. If not, adjust the consistency of the dynamic fuzzy judgment matrix until the consistency requirement is met. If so, determine that the dynamic fuzzy judgment matrix meets the consistency requirement; Perform fuzzy comprehensive operation on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain fuzzy weights; Use the weighted average method to defuzzify the fuzzy weights to obtain clear weight values, and use the clear weight values as subjective weights.

[0013] As can be seen from the above description, using the dynamic analytic network process to quantitatively analyze the importance of indicators based on expert opinions and finally obtain the subjective weights of each indicator can effectively integrate the experience and subjective judgment of experts, obtain the subjective weights of indicators, and improve the accuracy of the evaluation of support power technical indicators, which is especially suitable for complex systems that are difficult to directly evaluate through data.

[0014] Furthermore, the calculation of the consistency index of the dynamic fuzzy judgment matrix includes: ; In the formula, CI represents the consistency index of the dynamic fuzzy judgment matrix, λ max represents the largest eigenvalue of the dynamic fuzzy judgment matrix, n represents the matrix dimension of the dynamic fuzzy judgment matrix; The calculation of the random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index includes: ; In the formula, CR represents the random consistency ratio of the dynamic fuzzy judgment matrix, RI represents the random consistency index.

[0015] As can be seen from the above description, by calculating the consistency index and the random consistency ratio to test the consistency of the fuzzy judgment matrix and verify the reliability of the dynamic fuzzy judgment matrix, the calculation accuracy of the subjective weights can be ensured.

[0016] Furthermore, the performing of fuzzy comprehensive operation on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain fuzzy weights includes: ; In the formula, W i represents the fuzzy weight, a in represents the matrix element in the i-th row and n-th column of the dynamic fuzzy judgment matrix that meets the consistency requirement, a jnrepresents the matrix element in the \(j\)-th row and \(n\)-th column of the dynamic fuzzy judgment matrix that meets the consistency requirement. represents the multiplication operation of fuzzy numbers; The defuzzification of the fuzzy weights using the weighted average method to obtain the clear weight values includes: ; In the formula, W m represents the clear weight value, l represents the overall lowest evaluation value of the relative importance of all index pairs, m represents the overall medium evaluation value of the relative importance of all index pairs, u represents the overall highest evaluation value of the relative importance of all index pairs.

[0017] As can be seen from the above description, after performing fuzzy comprehensive operation on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain the fuzzy weights, the weighted average method is used to defuzzify the fuzzy weights to obtain the clear weight values, and then the subjective weights can be obtained. Fuzzy comprehensive operation can handle uncertainties and is more flexible, ensuring the scientificity and effectiveness of the calculation of subjective weights.

[0018] Furthermore, the calculation of the objective weight of each index using the weight fusion method based on entropy method, machine learning algorithm and grey relational analysis includes: Construct the original data matrix according to each index of each supporting power supply, and perform standardization processing on the original data matrix to obtain the standardized data matrix; Calculate the proportion of each index based on the standardized data matrix, and calculate the information entropy of each index according to the proportion of each index; Calculate the difference coefficient of each index according to the information entropy of each index, and calculate the first weight according to the difference coefficient of each index; Use the machine learning algorithm to calculate the second weight based on the standardized data matrix; Determine the reference sequence as the ideal value of each index and the comparison sequence as the values of each supporting power supply on each index; Calculate the grey correlation coefficient according to the reference sequence and the comparison sequence; Calculate the grey correlation degree of each index according to the grey correlation coefficient, and calculate the third weight according to the grey correlation degree of each index; Perform weighted fusion on the first weight, the second weight and the third weight to obtain the objective weight of each index.

[0019] As can be seen from the above description, after calculating the weights of each index using the entropy value method, machine learning algorithms, and grey relational analysis respectively, the first weight, the second weight, and the third weight are weighted and fused to obtain the objective weight of each index, which can extract the inherent importance from the objective data of the index, and combine the results of different calculation methods for weight fusion to ensure the objectivity and scientificity of the objective weight.

[0020] Further, calculating the first weight according to the coefficient of variation of each index includes: ; In the formula, represents the first weight, d j represents the coefficient of variation of index j, and n represents the total number of indexes; Calculating the third weight according to the grey relational degree of each index includes: ; In the formula, represents the third weight, r j represents the grey relational degree of index j.

[0021] As can be seen from the above description, calculating the first weight according to the coefficient of variation of each index, using the entropy value method to analyze the dispersion degree of index data, assigns higher weights to indexes with greater volatility, while calculating the third weight according to the grey relational degree of each index can accurately evaluate the correlation degree of each index with the ideal value, thus improving the accuracy of objective weight calculation.

[0022] Further, dynamically fusing the subjective weight of each index with the objective weight of each index to obtain the comprehensive weight of each index includes: Dynamically adjusting the proportion factor of the subjective weight; Fusing according to the proportion factor of the subjective weight, the subjective weight of each index, and the objective weight of each index to obtain the initial comprehensive weight of each index; Normalizing the initial comprehensive weight of each index to obtain the comprehensive weight of each index.

[0023] As can be seen from the above description, by dynamically adjusting the fusion ratio of the weights, that is, the proportion factor of the subjective weight, dynamic weight allocation is realized, ensuring the flexibility and scientificity of the comprehensive weight.

[0024] Further, fusing according to the proportion factor of the subjective weight, the subjective weight of each index, and the objective weight of each index to obtain the initial comprehensive weight of each index includes: ; In the formula, W final represents the initial comprehensive weight of the index, represents the subjective weight proportion factor, W m represents the subjective weight of the index, W t represents the objective weight of the index.

[0025] As can be seen from the above description, by fusing the subjective weight proportion factor, the subjective weight of each index, and the objective weight of each index, the initial comprehensive weight of each index is obtained, which can reduce the comprehensive evaluation error and make the weight distribution more reasonable.

[0026] Furthermore, the use of the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each index of the support power supply, and the obtained technical index evaluation results of multiple support power supplies include: Construct an original data matrix according to each index of each support power supply; Determine the positive ideal solution and the negative ideal solution based on the original data matrix; Use a dynamic adjustment factor to correct the positive ideal solution and the negative ideal solution respectively to obtain the corrected positive ideal solution and the corrected negative ideal solution; Perform standardization processing on the original data matrix to obtain a standardized data matrix; Calculate the first Euclidean distance between each support power supply and the corrected positive ideal solution based on the standardized data matrix, the comprehensive weight of each index, and the corrected positive ideal solution; Calculate the second Euclidean distance between each support power supply and the corrected negative ideal solution based on the standardized data matrix, the comprehensive weight of each index, and the corrected negative ideal solution; Calculate the comprehensive score of each support power supply according to the first Euclidean distance and the second Euclidean distance; Rank multiple support power supplies according to the comprehensive score of each support power supply to obtain the technical index evaluation results of multiple support power supplies.

[0027] As can be seen from the above description, introducing a dynamic adjustment factor to improve the TOPSIS method, using the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each index of the support power supply, and obtaining the technical index evaluation results of multiple support power supplies, which improves the flexibility and adaptability of the comprehensive evaluation system and is applicable to dynamic multi-dimensional scenarios.

[0028] Please refer to Figure 2, another embodiment of the present invention provides a technical index evaluation system for a supporting power supply for a new power system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, each step in the above-mentioned technical index evaluation method for a supporting power supply for a new power system is implemented.

[0029] The above-mentioned technical index evaluation method and system for a supporting power supply for a new power system of the present invention can be applied to a new power system, which will be described below through specific embodiments: Please refer to Figure 1 , Figure 3 and Figure 4 , the first embodiment of the present invention is: A technical index evaluation method for a supporting power supply for a new power system, including the steps of: S1. Construct the technical indexes of each supporting power supply, and the technical indexes include a clean and low-carbon first-level index, a safe and abundant first-level index, an economic and efficient first-level index, a supply-demand coordination first-level index, and a flexible and intelligent first-level index.

[0030] Among them, as Figure 3 shown, the clean and low-carbon first-level index includes a carbon emission second-level index and an environmental protection second-level index. The carbon emission second-level index includes a carbon emission third-level index, and the environmental protection second-level index includes a pollutant emission third-level index and an environmental impact degree third-level index.

[0031] The safe and abundant first-level index includes an electric power second-level index and an electric quantity second-level index. The electric power second-level index includes an effective capacity third-level index and an electric power reliability third-level index, and the electric quantity second-level index includes a generable electric quantity third-level index and an actual generated electric quantity third-level index.

[0032] The economic and efficient first-level index includes an efficiency second-level index and an economy second-level index. The efficiency second-level index includes an energy conversion efficiency third-level index, and the economy second-level index includes a power generation cost third-level index.

[0033] The supply-demand coordination first-level index includes a peak regulation second-level index and a frequency modulation second-level index. The peak regulation second-level index includes an adjustment range third-level index and an adjustment rate third-level index, and the frequency modulation second-level index includes a speed change rate third-level index and a rapidity third-level index.

[0034] The flexible and intelligent first-level index includes an inertia second-level index and an intelligentization second-level index. The inertia second-level index includes a moment of inertia third-level index and a recovery time third-level index, and the intelligentization second-level index includes an intelligentization level third-level index.

[0035] Each index in the subsequent steps is each third-level index.

[0036] The present invention constructs five first-level indicators of clean, low-carbon, safe, abundant, economic, efficient, supply-demand coordination, and flexible intelligence, which are refined into a multi-level secondary and tertiary indicator system, comprehensively covering the performance of each dimension of the supporting power supply, and forming an indicator system with scientificity, comprehensiveness, and dynamic adaptability.

[0037] S2. Determine the subjective weight of each indicator using the dynamic analytic network process, and calculate the objective weight of each indicator using a weight fusion method based on the entropy method, machine learning algorithms, and grey relational analysis, as Figure 4 shown, specifically including S2.1 - S2.16: S2.1. Obtain the lowest possible value, medium possible value, and highest possible value of the relative importance of each indicator to another indicator determined by experts.

[0038] S2.2. Determine the triangular fuzzy judgment matrix according to the lowest possible value, medium possible value, and highest possible value of the relative importance of each indicator to another indicator.

[0039] Specifically, obtain triangular fuzzy numbers according to the lowest possible value, medium possible value, and highest possible value of the relative importance of each indicator to another indicator, and construct a triangular fuzzy judgment matrix according to the triangular fuzzy numbers.

[0040] Among them, the triangular fuzzy numbers can well reflect the uncertainty of experts and have the characteristics of simplicity and intuitiveness. The triangular fuzzy numbers are: ; In the formula, a ij represents a triangular fuzzy number, l ij represents the lowest possible value of the relative importance of indicator i to indicator j, m ij represents the medium possible value of the relative importance of indicator i to indicator j, u ij represents the highest possible value of the relative importance of indicator i to indicator j.

[0041] The membership function of the triangular fuzzy number is: ; In the formula, x represents the indicator input value.

[0042] The triangular fuzzy judgment matrix is: ; ; In the formula, A represents the triangular fuzzy judgment matrix, and n represents the total number of indicators.

[0043] S2.3. Improve the triangular fuzzy judgment matrix using a dynamic adjustment factor to obtain a dynamic fuzzy judgment matrix.

[0044] Among them, the elements of the dynamic fuzzy judgment matrix of the dynamic fuzzy judgment matrix are: ; In the formula, A ij represents the element of the dynamic fuzzy judgment matrix, R ij represents the original correlation degree, which is calculated from expert scoring or historical data. λ represents the dynamic adjustment factor, which is related to the change range of real-time data.

[0045] The traditional Analytic Hierarchy Process (AHP) relies on expert scoring to generate a fixed fuzzy judgment matrix, and the weight calculation lacks flexibility and is difficult to adapt to dynamic data scenarios. In the present invention, a dynamic correlation degree adjustment mechanism is introduced in constructing the triangular fuzzy judgment matrix through the dynamic network analytic hierarchy process. When the data fluctuates greatly, the dynamic adjustment factor λ can improve the sensitivity of some weights. The dynamic fuzzy judgment matrix generated after the dynamic adjustment of the correlation degree can better reflect the weight changes in the real-time scenario.

[0046] In an alternative embodiment, it further includes: performing consistency adjustment on the dynamic fuzzy judgment matrix to ensure that the matrix meets the consistency requirements.

[0047] S2.4. Determine the membership function.

[0048] Specifically, the traditional triangular fuzzy number may not be able to well reflect the complexity of fuzziness in some scenarios. Therefore, the Gaussian fuzzy number is introduced, and its membership function is: ; In the formula, represents the membership function of the Gaussian fuzzy number, m represents the mean of the Gaussian fuzzy number, σ represents the standard deviation, which is dynamically adjusted according to the dispersion degree of the index weights, x represents the index input value, which is used to calculate its corresponding membership degree.

[0049] Here, the membership function of the Gaussian fuzzy number is used to replace the membership function of the original triangular fuzzy number. The membership function of the triangular fuzzy number is linearly increasing and decreasing, while the membership function of the Gaussian fuzzy number provides a smooth curve, which can more accurately express fuzziness and can perform smoother processing on the data.

[0050] S2.5. Calculate the consistency index of the dynamic fuzzy judgment matrix, and calculate the random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index.

[0051] Among them, the calculation of the consistency index of the dynamic fuzzy judgment matrix includes: ; In the formula, CI represents the consistency index of the dynamic fuzzy judgment matrix, λ max represents the maximum eigenvalue of the dynamic fuzzy judgment matrix, n represents the matrix dimension of the dynamic fuzzy judgment matrix; The calculation of the random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index includes: ; In the formula, CR represents the random consistency ratio of the dynamic fuzzy judgment matrix, RI represents the random consistency index, and its value is related to the matrix dimension n related.

[0052] S2.6. Judge whether the random consistency ratio is less than the preset value. If not, perform consistency adjustment on the dynamic fuzzy judgment matrix until the consistency requirement is met. If so, determine that the dynamic fuzzy judgment matrix meets the consistency requirement.

[0053] In an optional implementation manner, the preset value is 0.1.

[0054] Among them, the consistency adjustment of the dynamic fuzzy judgment matrix includes: By adjusting the upper and lower bounds a ij of the triangular fuzzy number l ij and u ij reconstruct the dynamic fuzzy judgment matrix or consult experts again to complete the consistency adjustment.

[0055] S2.7. Perform fuzzy comprehensive operation on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain the fuzzy weight, specifically: ; In the formula, W i represents the fuzzy weight, a in represents the matrix element in the i-th row and n-th column of the dynamic fuzzy judgment matrix that meets the consistency requirement, a jnIt represents the matrix element in the \(j\)-th row and \(n\)-th column of the dynamic fuzzy judgment matrix that meets the consistency requirements. It represents the multiplication operation of fuzzy numbers.

[0056] S2.8. Use the weighted average method to defuzzify the fuzzy weights to obtain the crisp weight values, and take the crisp weight values as the subjective weights.

[0057] Among them, the use of the weighted average method to defuzzify the fuzzy weights to obtain the crisp weight values includes: ; In the formula, W m represents the crisp weight value, l represents the overall lowest evaluation value of the relative importance of all index pairs, m represents the overall medium evaluation value of the relative importance of all index pairs, u represents the overall highest evaluation value of the relative importance of all index pairs.

[0058] Adopt the dynamic analytic network process to fuzzify the subjective judgments of experts, generate a fuzzy judgment matrix, and calculate the subjective weights of each index through the fuzzy weighting method. This method effectively solves the problems of strong subjectivity and insufficient fuzziness in the traditional analytic hierarchy process, and ensures the scientificity of weight calculation through consistency verification.

[0059] S2.9. Construct an original data matrix according to each index of each supporting power supply, and perform standardization processing on the original data matrix to obtain a standardized data matrix.

[0060] Among them, the construction of the original data matrix according to each index of each supporting power supply X = x ij includes: ; In the formula, m represents the total number of supporting power supplies, n represents the total number of indexes, x ij represents the i -th value of the j -th supporting power supply under the

[0061] The standardization processing of the original data matrix to obtain the standardized data matrix Y = y ij includes: If the index is a positive index (i.e., the larger the value, the better), then: ; If the index is a negative index (i.e., the smaller the value, the better), then: ; In the formula, y ij represents the matrix element in the standardized data matrix, y ij is between 0 and 1.

[0062] S2.10. Calculate the proportion of each index based on the standardized data matrix, and calculate the information entropy of each index according to the proportion of each index.

[0063] Among them, the calculation of the proportion of each index based on the standardized data matrix includes: ; In the formula, p ij represents the proportion of the j th index in the i th supporting power supply.

[0064] The calculation of the information entropy of each index according to the proportion of each index includes: ; In the formula, e j represents the information entropy of the j th index, k represents the normalization coefficient, , which is used to ensure that e j is within the range of [0,1].

[0065] S2.11. Calculate the coefficient of variation of each index according to the information entropy of each index, and calculate the first weight according to the coefficient of variation of each index; Among them, the calculation of the coefficient of variation of each index according to the information entropy of each index includes: ; In the formula, d j represents the coefficient of variation of the j th index, and the coefficient of variation is used to measure the dispersion degree of each index.

[0066] The calculation of the first weight according to the coefficient of variation of each index includes: ; In the formula, represents the first weight, and n represents the total number of indexes.

[0067] The entropy value method is a classical objective weight calculation method. By analyzing the dispersion degree of index data, higher weights are assigned to the indexes with greater volatility.

[0068] S2.12. Use the machine learning algorithm to calculate the second weight based on the standardized data matrix, specifically including S2.12.1 - S2.12.4: S2.12.1. Prepare data: Input data: The standardized data matrix Y = y ij ; Output data: The target variable Z (usually the comprehensive score of the scheme or the evaluation target value).

[0069] S2.12.2. Select a model: Use the random forest model or the XGBoost model, which are suitable for processing multi-dimensional features and non-linear relationships. The random forest model fits the data through multiple decision trees and evaluates the importance of each feature (index) in the splitting; the XGBoost model optimizes the fitting accuracy of the model by means of weighted boosting trees.

[0070] S2.12.3. Train the model: Use Y and Z to train the model to obtain the feature importance of each feature F j .

[0071] S2.12.4. Normalization processing: Normalize the feature importance to the second weight, specifically: ; In the formula, represents the second weight.

[0072] Use the machine learning algorithm to mine the contribution degree of the index to the evaluation result and learn the weight assignment from the data.

[0073] S2.13. Determine the reference sequence Y 0 = y 01 , y 02 , …, y 0n is the ideal value of each index and the comparison sequence Y i = y i1 , yi2 ,…, y in are the values of each of the said supporting power supplies in each index.

[0074] S2.14. Calculate the grey correlation coefficient according to the said reference sequence and the said comparison sequence, specifically: ; In the formula, represents the grey correlation coefficient, represents the discrimination coefficient. In an optional implementation manner, = 5.

[0075] S2.15. Calculate the grey correlation degree of each index according to the said grey correlation coefficient, and calculate the third weight according to the grey correlation degree of each index.

[0076] Among them, the calculation of the grey correlation degree of each index according to the said grey correlation coefficient includes: ; In the formula, r j represents the grey correlation degree of index j.

[0077] The calculation of the third weight according to the grey correlation degree of each index includes: ; In the formula, represents the third weight.

[0078] Grey correlation analysis is a calculation method based on the similarity of system behavior patterns, used to evaluate the correlation degree of each index with the ideal value.

[0079] S2.16. Perform weighted fusion on the said first weight, the said second weight and the said third weight to obtain the objective weight of each index, specifically: ; In the formula, W t represents the objective weight of the index, represents the first weight adjustment coefficient, represents the second weight adjustment coefficient, represents the third weight adjustment coefficient. The weight adjustment coefficient can be dynamically determined by an optimization algorithm.

[0080] An objective weight calculation method using multi-source data fusion combines the entropy method, machine learning feature importance analysis, and grey relational analysis to dynamically evaluate the volatility and contribution degree of index data and calculate the objective weights of each index. Through this method, the actual contributions of the indexes in different scenarios can be comprehensively reflected and the requirements of technological development can be met.

[0081] S3. Dynamically fuse the subjective weight of each index with the objective weight of each index to obtain the comprehensive weight of each index, as Figure 4 shown, specifically including S3.1 - S3.3: S3.1. Dynamically adjust the subjective weight proportion factor.

[0082] Among them, the subjective weight proportion factor is a dynamic parameter that can be adjusted according to the actual evaluation scenario or real-time data. For example, when expert experience is more important, increase the weight proportion of the subjective weight proportion factor and rely more on the subjective weight; when data-driven is more reliable, reduce the subjective weight proportion factor and rely more on the objective weight. The range of the subjective weight proportion factor is [0, 1].

[0083] In an alternative implementation, the dynamic adjustment of the subjective weight proportion factor includes: Set adjustment rules according to the characteristics of the evaluation system and real-time data fluctuations, and adjust the subjective weight proportion factor according to the adjustment rules.

[0084] For example, the adjustment rule is: if the volatility of the index data is large (such as a high standard deviation), then reduce the subjective weight proportion factor and rely more on the objective weight; if the subjective consistency between indexes is strong, then increase the subjective weight proportion factor and rely more on the subjective weight.

[0085] In an alternative implementation, the dynamic adjustment of the subjective weight proportion factor includes: Use an optimization algorithm (such as a genetic algorithm, particle swarm optimization algorithm) to dynamically optimize the subjective weight proportion factor and output the optimal subjective weight proportion factor to ensure the scientificity and adaptability of the subjective and objective weight ratios.

[0086] Specifically, the optimization objective is: ; where f() represents the error or consistency deviation of the comprehensive evaluation result.

[0087] The optimization constraint is: 0 ≤ β ≤ 1; Algorithm output: Output the optimal β value.

[0088] S3.2. Integrate the subjective weight ratio factor, the subjective weight of each indicator, and the objective weight of each indicator to obtain the initial comprehensive weight of each indicator. Specifically: ; In the formula, W final represents the initial comprehensive weight of the indicator, represents the subjective weight ratio factor.

[0089] S3.3. Normalize the initial comprehensive weight of each indicator to obtain the comprehensive weight of each indicator to ensure the effectiveness of the comprehensive weight. Specifically: ; In the formula, W final ′ represents the comprehensive weight of the indicator.

[0090] Combine the subjective weight and the objective weight, and realize the dynamic weight distribution by dynamically adjusting the fusion ratio of the weights, so as to ensure the adaptability and scientificity of the weights.

[0091] S4. Use the improved TOPSIS method to rank multiple support power supplies based on the comprehensive weight of each indicator of the support power supply, and obtain the technical index evaluation results of multiple support power supplies, specifically including S4.1 - S4.8: S4.1. Construct the original data matrix according to each indicator of each support power supply X = x ij .

[0092] S4.2. Determine the positive ideal solution and the negative ideal solution based on the original data matrix.

[0093] Among them, the positive ideal solution V + , representing the optimal value of each indicator, is: .

[0094] The negative ideal solution V - , representing the worst value of each indicator, is: .

[0095] S4.3. Use the dynamic adjustment factor to correct the positive ideal solution and the negative ideal solution respectively to obtain the corrected positive ideal solution and the corrected negative ideal solution to further adapt to the changes of the scenario. Specifically: ; ; In the formula, f j represents a dynamic adjustment factor, which can be dynamically calculated based on the volatility of the indicators or other weight adjustment logics. represents the corrected positive ideal solution, represents the corrected negative ideal solution.

[0096] S4.4. Standardize the original data matrix to obtain the standardized data matrix X ′ = x ij ′], to eliminate the dimensional differences of different indicators.

[0097] Specifically, if the indicator is a positive indicator, then: .

[0098] If the indicator is a negative indicator, then: .

[0099] Among them, all the data in the standardized data matrix X ′ = x ij ′] are between [0, 1].

[0100] S4.5. Calculate the first Euclidean distance between each supporting power supply and the corrected positive ideal solution based on the standardized data matrix, the comprehensive weight of each indicator, and the corrected positive ideal solution, specifically: ; In the formula, D i + represents the first Euclidean distance between a certain supporting power supply and the corrected positive ideal solution.

[0101] S4.6. Calculate the second Euclidean distance between each supporting power supply and the corrected negative ideal solution based on the standardized data matrix, the comprehensive weight of each indicator, and the corrected negative ideal solution, specifically: ; In the formula, D i - represents the second Euclidean distance between the supporting power supply i and the corrected negative ideal solution.

[0102] S4.7. Calculate the comprehensive score of each supporting power supply according to the first Euclidean distance and the second Euclidean distance, specifically: ; In the formula,C i Represents the comprehensive score of the supporting power supply i, with a value range between [0, 1]. C i The larger it is, the closer the solution is to the ideal solution and the higher the priority.

[0103] S4.8. Sort the multiple supporting power supplies according to the comprehensive score of each supporting power supply to obtain the technical index evaluation results of the multiple supporting power supplies.

[0104] Specifically, sort the multiple supporting power supplies according to the comprehensive score of each supporting power supply in descending numerical order to obtain the technical index evaluation results of the multiple supporting power supplies. The supporting power supply with the highest comprehensive score is the optimal solution.

[0105] The present invention constructs a technical index system with five dimensions of clean and low-carbon, safe and abundant, economic and efficient, supply-demand coordination, and flexible and intelligent, comprehensively covering the core characteristics of the supporting power supply, and improves the scientificity of the evaluation system through a weight calculation method combining subjective and objective. The multi-source data fusion method can dynamically adjust the weight and evaluation model to adapt to multi-scenario applications of technological progress and market demand. Using the dynamic network analytic hierarchy process and the weight fusion method based on the entropy method, machine learning algorithms, and grey relational analysis reduces the data processing and calculation complexity, improves the applicability of the model and the reliability of the results. At the same time, it enhances the dynamic adaptation ability of the evaluation model to complex multi-dimensional scenarios, providing scientific and reliable decision-making support for the planning and optimization of the new power system.

[0106] Please refer to Figure 2 , and the second embodiment of the present invention is as follows: A technical index evaluation system for a supporting power supply for a new power system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, each step in the technical index evaluation method for a supporting power supply for a new power system in Embodiment 1 is implemented.

[0107] In summary, the present invention provides a method and system for evaluating the technical indicators of a supporting power source for a new power system, constructs the technical indicators of each supporting power source, and the technical indicators include primary indicators of clean and low-carbon, primary indicators of safety and adequacy, primary indicators of economic efficiency, primary indicators of supply-demand coordination, and primary indicators of flexibility and intelligence. The comprehensive evaluation considering multi-dimensional characteristics comprehensively covers the performance of each dimension of the supporting power source. The subjective weight of each indicator is determined using the dynamic network analytic hierarchy process, effectively solving the problems of strong subjectivity and insufficient fuzziness in the traditional analytic hierarchy process. The objective weight of each indicator is calculated using a weight fusion method based on the entropy method, machine learning algorithms, and grey relational analysis, which can comprehensively reflect the actual contribution of the indicators in different scenarios and adapt to policy changes and technological development needs. The subjective weight and objective weight of each indicator are dynamically fused to obtain the comprehensive weight of each indicator, effectively balancing the influence of subjective weight and objective weight, improving the adaptability and scientificity of the comprehensive weight calculation. Finally, the improved TOPSIS method is used to rank multiple supporting power sources based on the comprehensive weight of each indicator of the supporting power source, and the technical indicator evaluation results of multiple supporting power sources are obtained, improving the dynamic adaptation ability of the evaluation model to complex multi-dimensional scenarios, reducing data processing and calculation complexity, thereby improving the comprehensiveness, flexibility, and simplicity of the technical indicator evaluation of the supporting power source; at the same time, a dynamic adjustment factor is introduced to improve the TOPSIS method, and the improved TOPSIS method is used to rank multiple supporting power sources based on the comprehensive weight of each indicator of the supporting power source, and the technical indicator evaluation results of multiple supporting power sources are obtained, improving the flexibility and adaptability of the comprehensive evaluation system and being applicable to dynamically changing multi-dimensional scenarios.

[0108] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent transformation made using the content of the specification and drawings of the present invention, or directly or indirectly applied in the relevant technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for evaluating technical indicators of supporting power supply for new power systems, characterized in that: Includes steps: Establish technical indicators for each supporting power source, including clean and low-carbon first-level indicators, safe and sufficient first-level indicators, economic and efficient first-level indicators, supply and demand coordination first-level indicators, and flexible and intelligent first-level indicators; The subjective weight of each indicator is determined using the dynamic network analytic hierarchy process, and the objective weight of each indicator is calculated using a weight fusion method based on entropy method, machine learning algorithm and grey relational analysis; Dynamically integrating the subjective weight of each indicator with the objective weight of each indicator to obtain a comprehensive weight of each indicator; An improved TOPSIS method is used to sort the plurality of supporting power sources based on the comprehensive weight of each indicator of the supporting power source, so as to obtain technical indicator evaluation results of the plurality of supporting power sources.

2. According to claim 1, a method for evaluating technical indicators of supporting power supply for a new power system is characterized in that: The method of using the dynamic network analytic hierarchy process to determine the subjective weight of each indicator includes: Obtain the lowest possible, medium possible, and highest possible values ​​of the relative importance of each indicator to the other indicator as determined by experts; Determine a triangular fuzzy judgment matrix according to the lowest possible value, the middle possible value and the highest possible value of the relative importance of each indicator to another indicator; Using a dynamic adjustment factor to improve the triangular fuzzy judgment matrix to obtain a dynamic fuzzy judgment matrix; Determine the membership function; Calculating a consistency index of the dynamic fuzzy judgment matrix, and calculating a random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index; Determine whether the random consistency ratio is less than a preset value, if not, adjust the consistency of the dynamic fuzzy judgment matrix until the consistency requirement is met, if yes, determine that the dynamic fuzzy judgment matrix meets the consistency requirement; Performing fuzzy comprehensive operation on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain fuzzy weights; The fuzzy weight is defuzzified using a weighted average method to obtain a clear weight value, and the clear weight value is used as a subjective weight.

3. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 2 is characterized in that: The calculation of the consistency index of the dynamic fuzzy judgment matrix includes: ； In the formula, CI represents the consistency index of the dynamic fuzzy judgment matrix, λ max represents the maximum eigenvalue of the dynamic fuzzy judgment matrix, n Represents the matrix dimension of the dynamic fuzzy judgment matrix; The step of calculating the random consistency ratio of the dynamic fuzzy judgment matrix according to the consistency index comprises: ； In the formula, CR represents the random consistency ratio of the dynamic fuzzy judgment matrix, RI Represents the random consistency index.

4. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 2, characterized in that: The fuzzy comprehensive operation is performed on the dynamic fuzzy judgment matrix that meets the consistency requirement to obtain the fuzzy weights, which includes: ； In the formula, W i represents the fuzzy weight, a in represents the matrix element in the i-th row and n-th column of the dynamic fuzzy judgment matrix that meets the consistency requirements, a jn represents the matrix element in the jth row and nth column of the dynamic fuzzy judgment matrix that meets the consistency requirements, Represents the multiplication operation of fuzzy numbers; The step of defuzzifying the fuzzy weight using a weighted average method to obtain a clear weight value comprises: ； In the formula, W m Indicates the clear weight value, l The overall lowest evaluation value representing the relative importance of all pairs of indicators, m The overall median assessment of the relative importance of all pairs of indicators, u The overall highest evaluation value representing the relative importance of all pairs of indicators.

5. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 1 is characterized in that: The objective weight of each indicator is calculated by using a weight fusion method based on entropy method, machine learning algorithm and grey relational analysis, including: Constructing an original data matrix according to each indicator of each supporting power source, and performing standardization processing on the original data matrix to obtain a standardized data matrix; Calculating the weight of each indicator based on the standardized data matrix, and calculating the information entropy of each indicator according to the weight of each indicator; Calculating a difference coefficient of each indicator according to the information entropy of each indicator, and calculating a first weight according to the difference coefficient of each indicator; Calculate a second weight based on the standardized data matrix using a machine learning algorithm; Determine the reference sequence as the ideal value of each indicator and the comparison sequence as the value of each supporting power source on each indicator; Calculate the grey correlation coefficient according to the reference sequence and the comparison sequence; Calculating the grey relational degree of each indicator according to the grey relational coefficient, and calculating the third weight according to the grey relational degree of each indicator; The first weight, the second weight and the third weight are weighted and fused to obtain an objective weight of each indicator.

6. A method for evaluating technical indicators of supporting power supply for a new power system according to claim 5, characterized in that: Calculating the first weight according to the difference coefficient of each indicator includes: ； In the formula, represents the first weight, d j represents the coefficient of variation of indicator j, and n represents the total number of indicators; Calculating the third weight according to the grey relational degree of each indicator includes: ； In the formula, represents the third weight, r j Represents the grey relational degree of index j.

7. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 1, characterized in that: The subjective weight of each indicator is dynamically integrated with the objective weight of each indicator to obtain the comprehensive weight of each indicator, including: Dynamically adjust the subjective weight ratio factor; The initial comprehensive weight of each indicator is obtained by fusing the subjective weight ratio factor, the subjective weight of each indicator and the objective weight of each indicator; The initial comprehensive weight of each indicator is normalized to obtain the comprehensive weight of each indicator.

8. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 7, characterized in that: The initial comprehensive weight of each indicator is obtained by fusing the subjective weight ratio factor, the subjective weight of each indicator and the objective weight of each indicator, including: ； In the formula, W final represents the initial comprehensive weight of the indicator, represents the subjective weight factor, W m represents the subjective weight of the indicator, W t Indicates the objective weight of the indicator.

9. The method for evaluating technical indicators of supporting power supply for a new power system according to claim 1, characterized in that: The improved TOPSIS method is used to sort the plurality of supporting power sources based on the comprehensive weight of each indicator of the supporting power source to obtain the technical indicator evaluation results of the plurality of supporting power sources, including: Construct a raw data matrix based on each indicator of each supporting power source; Determining a positive ideal solution and a negative ideal solution based on the original data matrix; Using a dynamic adjustment factor to correct the positive ideal solution and the negative ideal solution respectively, to obtain a corrected positive ideal solution and a corrected negative ideal solution; Performing standardization processing on the original data matrix to obtain a standardized data matrix; Calculating a first Euclidean distance between each supporting power source and the modified positive ideal solution based on the standardized data matrix, the comprehensive weight of each indicator and the modified positive ideal solution; Calculating a second Euclidean distance between each supporting power source and the modified negative ideal solution based on the standardized data matrix, the comprehensive weight of each indicator and the modified negative ideal solution; Calculate a comprehensive score of each supporting power source according to the first Euclidean distance and the second Euclidean distance; The plurality of supporting power sources are sorted according to the comprehensive score of each supporting power source to obtain technical indicator evaluation results of the plurality of supporting power sources.

10. A supporting power supply technical indicator evaluation system for a new power system, 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, each step of a method for evaluating technical indicators of supporting power sources for a new power system according to any one of claims 1 to 9 is implemented.

Images