A substation carbon emission early warning method and terminal

Abstract

The application discloses a substation carbon emission early warning method and a terminal, experts evaluate the mutual influence degree between different target indexes based on relevant experience, thereby obtaining subjective evaluation opinions, and the subjective evaluation opinions are quantified into subjective weights through an evaluation language set, so that the internal connection between the target indexes in a complex system is reflected in this way. Meanwhile, the subjective weights and objective weights determined by an entropy weight method are combined to determine the comprehensive weights of the target indexes, the subjective opinions of the experts are combined with objective results of scientific analysis, and the objectivity and reliability of index weighting are improved. Considering the uncertainty and incompleteness of the target index data, the grey correlation analysis is introduced based on the correlation degree between different target indexes, so that a decision matrix between the indexes is determined, data blanks of the target indexes are effectively filled, and the early warning interval is determined through the Raimi criterion, so that classification errors of the early warning area caused by subjective judgment can be avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission early warning, and in particular to a method and terminal for early warning of carbon emissions from a substation. Background Art

[0002] Carbon emission early warning for substations is a key component of environmental protection and carbon emission management, crucial for reducing greenhouse gas emissions and achieving sustainable development. In the context of low-carbon development, establishing an accurate and reliable early warning mechanism can provide theoretical guidance for optimizing substation energy structures and achieving a green, low-carbon transition. However, the continuous advancement of intelligent substations and the increasing coupling between various energy sources within substations have posed significant challenges to carbon emission early warning.

[0003] In substation carbon emission early warning, indicator empowerment and data integration are two aspects that currently face difficulties.

[0004] When it comes to indicator weighting, correctly measuring the importance of each indicator and determining appropriate weightings is a challenging task. Different indicators interact with each other, and the degree of mutual influence between them must be comprehensively considered to accurately assess their importance. Expertise and experience in relevant fields are also indispensable to the weighting process. Substation carbon emissions involve multiple fields, including power systems, environmental science, and energy management, and the accuracy of indicator weighting often relies on the judgment and understanding of domain experts. A lack of relevant expertise or the involvement of domain experts can lead to inaccurate indicator weighting. In practice, the lack of clear standards and methods makes it difficult to accurately determine indicator weights, which can also affect the reliability of early warning results.

[0005] Carbon emission early warnings require comprehensive consideration of multiple parameters or indicators, often from disparate data sources. The diversity of data sources and inconsistent data formats make data integration difficult. Furthermore, data quality issues can impact the accuracy of early warning results. Data loss, data anomalies, and incompatibilities between data sources can negatively impact early warning results. Therefore, ensuring data consistency and reliability is crucial during the data integration process. Traditional approaches often rely solely on a single indicator as the basis for early warnings, overlooking the importance of comprehensive consideration of multiple indicators. This single-indicator approach often fails to accurately analyze the overall carbon emissions profile of a substation, resulting in incomplete and inaccurate early warning results.

[0006] Therefore, traditional methods lack scientificity and flexibility in indicator empowerment and data integration, and are unable to analyze and judge the complex relationships between early warning indicators. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide an early warning method and terminal for carbon emissions from substations, which combines the subjective opinions of experts with the objective results of scientific analysis to improve the objectivity and reliability of indicator weighting. At the same time, the connections and patterns between multiple indicators can be determined based on abnormal data to improve the accuracy of early warning.

[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0009] A method for early warning of carbon emissions from a substation, comprising:

[0010] Determine target indicators that affect substation carbon emissions based on the PSR theoretical model;

[0011] Obtain subjective evaluation opinions of different experts on the degree of impact between the target indicators;

[0012] Converting the subjective evaluation opinions into subjective weights through a preset evaluation language set;

[0013] Determine the objective weight of each target indicator through the entropy weight method;

[0014] Combining and weighting the subjective weight and the objective weight to obtain a comprehensive weight between the target indicators;

[0015] Performing grey correlation analysis on the target indicators to obtain a decision matrix, and obtaining a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix;

[0016] The warning interval of the warning value output by the substation carbon emission warning indicator model is determined by the Raida criterion, and the warning interval is classified into warning levels.

[0017] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0018] A substation carbon emission early warning terminal includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, each step of the above-mentioned substation carbon emission early warning method is implemented.

[0019] The beneficial effects of the present invention are as follows: based on the PSR (Pressure-State-Response) theoretical framework, target indicators affecting substation carbon emissions are determined. Experts, based on their professional knowledge and relevant experience, evaluate the degree of mutual influence between different target indicators, thereby obtaining subjective evaluation opinions. These subjective evaluation opinions, in text form, are then quantified into numerical subjective weights using an evaluation language set. This reflects the inherent connections between target indicators in a complex system and provides a more comprehensive assessment of the importance of each target indicator in the indicator system. Furthermore, the objective weights of the target indicators are determined based on the entropy weight method, and a combined weighting is performed based on the subjective and objective weights to determine the comprehensive weight of each target indicator. This combines the subjective opinions of experts with the objective results of scientific analysis, thereby improving the objectivity and reliability of the indicator weighting. Furthermore, considering the uncertainty and incompleteness of target indicator data, gray correlation analysis is introduced based on the degree of correlation between different target indicators to determine the decision matrix between the indicators, effectively helping to fill the data gaps of the target indicators. Furthermore, the early warning interval is determined using the Raida criterion, which can avoid the classification errors caused by subjective judgment in the early warning area, effectively improving the accuracy of the substation carbon emission early warning indicator model. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flowchart of the steps of a method for early warning of carbon emissions from a substation provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of a carbon emission early warning indicator system for a substation provided by an embodiment of the present invention;

[0022] Figure 3 A flowchart of another step of a method for early warning of carbon emissions from a substation provided by an embodiment of the present invention;

[0023] Figure 4 A schematic structural diagram of a substation carbon emission early warning terminal provided by an embodiment of the present invention;

[0024] Description of labels:

[0025] 300. An early warning terminal for carbon emissions from a substation; 301. Memory; 302. Processor. DETAILED DESCRIPTION

[0026] To illustrate the technical content, achieved objectives and effects of the present invention in detail, the following description is given in conjunction with the embodiments and accompanying drawings.

[0027] Please refer to Figure 1 The embodiment of the present invention provides a method for early warning of carbon emissions from a substation, comprising:

[0028] Determine target indicators that affect substation carbon emissions based on the PSR theoretical model;

[0029] Obtain subjective evaluation opinions of different experts on the degree of impact between the target indicators;

[0030] Converting the subjective evaluation opinions into subjective weights through a preset evaluation language set;

[0031] Determine the objective weight of each target indicator through the entropy weight method;

[0032] Combining and weighting the subjective weight and the objective weight to obtain a comprehensive weight between the target indicators;

[0033] Performing grey correlation analysis on the target indicators to obtain a decision matrix, and obtaining a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix;

[0034] The warning interval of the warning value output by the substation carbon emission warning indicator model is determined by the Raida criterion, and the warning interval is classified into warning levels.

[0035] As can be seen from the above description, the beneficial effects of the present invention are as follows: Target indicators influencing substation carbon emissions are determined based on the PSR theoretical model. Experts, based on their professional knowledge and relevant experience, assess the degree of mutual influence between different target indicators, thereby obtaining subjective evaluation opinions. These subjective evaluation opinions, expressed in text form, are then quantified into numerical subjective weights using an evaluation language set. This approach reflects the inherent connections between target indicators in a complex system and provides a more comprehensive assessment of the importance of each target indicator within the indicator system. Furthermore, objective weights for target indicators are determined based on the entropy weight method, and a combined weighting is performed based on the subjective and objective weights to determine the overall weight for each target indicator. This combines the subjective opinions of experts with the objective results of scientific analysis, improving the objectivity and reliability of indicator weighting. Furthermore, considering the uncertainty and incompleteness of target indicator data, gray correlation analysis is introduced based on the degree of correlation between different target indicators to determine the decision matrix between the indicators, effectively helping to fill the data gaps for the target indicators. Furthermore, the early warning interval is determined using the Raida criterion, which avoids classification errors caused by subjective judgment within the early warning area and effectively improves the accuracy of the substation carbon emission early warning indicator model.

[0036] Furthermore, converting the subjective evaluation opinions into subjective weights using a preset evaluation language set includes:

[0037] Converting the subjective evaluation opinions into intuitive fuzzy numbers through a preset evaluation language set;

[0038] Calculating the comprehensive impact value of each target indicator according to the intuitionistic fuzzy number;

[0039] The subjective weight of each target indicator is determined according to the comprehensive impact value.

[0040] As can be seen from the above description, intuitionistic fuzzy numbers are used to quantify the degree of influence between different target indicators. Based on human subjective logical thinking and expression, this approach directly quantifies the expert's evaluation language into intuitionistic fuzzy numbers, which are then converted into subjective weights. This fully leverages expert experience and comprehensively considers the degree of coupling between different target indicators, improving the pertinence and accuracy of indicator data analysis. Furthermore, compared to directly assigning weights by experts, this approach of quantifying evaluation language into intuitionistic fuzzy numbers makes the analysis results more interpretable and scientific, thanks to the experts' subjective judgment and experience.

[0041] Furthermore, the determining of the objective weight of each target indicator by the entropy weight method includes:

[0042] Standardizing the sample data of each target indicator to obtain a standard factor value;

[0043] Calculating the proportion of the target indicator in the sample data according to the standard factor value;

[0044] Calculate the entropy value of each target indicator according to the proportion;

[0045] The objective weight of each target indicator is determined according to the entropy value.

[0046] As can be seen from the above description, the entropy weight method uses information entropy to calculate the entropy weight of each target indicator based on the degree of variation of different target indicators. The entropy weight is then used to modify the weight of each target indicator to obtain objective indicator weights. Based on objective weights, the bias and subjectivity that may exist in subjective weights are eliminated to improve the objectivity and scientific nature of weight allocation.

[0047] Furthermore, the combining and weighting according to the subjective weight and the objective weight to obtain the comprehensive weight between the target indicators includes:

[0048] Performing combined weighting according to the subjective weight and the objective weight to obtain a combined weight;

[0049] Calculating the combined weights according to the principle of the sum of squared deviations, and constructing an objective function with the maximum sum of squared deviations as the goal;

[0050] Solve the objective function and determine the comprehensive weights between the objective indicators based on the solution results.

[0051] As can be seen from the above description, the sum of squared differences (SSD) is a method used to calculate differences between data. In combination weighting, SSD can minimize differences between data. In the combination weighting process, improperly selecting weights for different target indicators can result in minimal differences in carbon emission warning values ​​across different scenarios, hindering accurate warnings in different scenarios. Therefore, when assigning weights, it is necessary to construct an objective function that maximizes the sum of squared differences. This ensures that carbon emission warning values ​​across different scenarios vary significantly, improving warning accuracy.

[0052] Furthermore, performing grey correlation analysis on the target indicators to obtain a decision matrix, and obtaining a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix includes:

[0053] Determine the original data matrix and the reference sequence according to the target indicator;

[0054] Performing dimensionless processing on the original data matrix to obtain a standard data matrix;

[0055] Calculate the grey relational coefficient of each target indicator according to the standard data matrix and the reference sequence;

[0056] Determine a decision matrix according to the grey correlation coefficient;

[0057] A substation carbon emission early warning indicator model is obtained according to the comprehensive weight and the decision matrix.

[0058] From the above description, it can be seen that grey correlation analysis can effectively process incomplete information in target indicators, quantify the degree of correlation between target indicators, improve the accuracy of the early warning system, and by calculating the grey correlation coefficient, it can intuitively obtain key indicators with greater impact on carbon emissions, help optimize the early warning model, and provide decision makers with accurate data reference.

[0059] Furthermore, the comprehensive impact value includes a first impact value that affects other target indicators and a second impact value that is affected by other target indicators;

[0060] Calculating the comprehensive impact value of each target indicator according to the intuitionistic fuzzy number includes:

[0061] Aggregate the intuitionistic fuzzy numbers of different experts for the same target indicator to obtain the aggregated intuitionistic fuzzy number:

[0062]

[0063] Among them, fij represents the aggregated intuitionistic fuzzy number, represents the intuitionistic fuzzy number of the kth expert on the target indicator, represents the expected value of the intuitionistic fuzzy number, Indicates the fuzziness degree of the intuitionistic fuzzy number, represents the confidence of the intuitionistic fuzzy number, and λk represents the aggregation weight of the kth expert;

[0064] Obtaining a direct influence matrix according to the aggregated intuitionistic fuzzy number, and defuzzifying and standardizing the direct influence matrix to obtain a standard influence matrix;

[0065] Determine the comprehensive impact matrix based on the standard impact matrix:

[0066] T=[t ij ] n×n =Y(IY) -1 ;

[0067] Among them, T represents the comprehensive impact matrix, t ij represents the impact value of target indicator i on target indicator j, Y represents the standard impact matrix, and I represents the unit matrix;

[0068] Determine the first impact value and the second impact value according to the comprehensive impact matrix:

[0069]

[0070]

[0071] Among them, R i represents the first impact value of the j-th target indicator, C j Represents the second impact value of the j-th target indicator.

[0072] As can be seen from the above description, the comprehensive impact value, consisting of both the first and second impact values, comprehensively considers the direct and indirect effects between different target indicators, thereby more accurately determining the combined weights of different indicators. This approach fully leverages the expertise and relevant experience of experts, comprehensively considers the degree of coupling between different target indicators, and thus improves the accuracy of indicator data analysis and early warning.

[0073] Furthermore, determining the subjective weight of each target indicator according to the comprehensive impact value includes:

[0074] The original weight of each target indicator is calculated according to the first impact value and the second impact value, and the original weight is normalized to obtain a subjective weight:

[0075]

[0076]

[0077] Among them, w sj represents the subjective weight of the jth target indicator, w' sj represents the original weight of the j-th target indicator.

[0078] From the above description, it can be seen that the subjective weight is determined based on the impact value between different target indicators, which scientifically quantifies the degree of influence between different indicators, thereby providing a more scientific, objective and comprehensive evaluation effect on the importance of each indicator in carbon emission early warning.

[0079] Furthermore, the combining weighting according to the subjective weight and the objective weight to obtain the combined weight includes:

[0080] W c =θ1W1+θ2W2;

[0081] Among them, W c Represents the combined weight of all target indicators, W1 represents the subjective weight of all target indicators, and W2 represents the objective weight of all target indicators; θ1 and θ2 represent the combination coefficients, and θ1 and θ2 satisfy the constraints

[0082] The step of calculating the combined weights according to the principle of the sum of squared deviations and constructing an objective function with the maximum sum of squared deviations as the goal includes:

[0083]

[0084] Among them, J(ω) represents the objective function, S i (W c ) indicates W c The sum of squares of the deviations between the i-th combined weight method and other methods, ω j 、 and W c The elements in b ij Indicates the sample data value of the target indicator.

[0085] From the above description, it can be seen that combined weighting based on the sum of squared deviations can combine the subjective opinions of experts and the comprehensive influence of different target indicators, thereby making up for the defects of objective weighting and allocating the weight values ​​of different target indicators more accurately and reasonably.

[0086] Furthermore, the calculating of the grey relational coefficient of each target indicator according to the standard data matrix and the reference sequence includes:

[0087]

[0088] Among them, ε i (k) represents the grey correlation coefficient of the kth target indicator, x' o (k) represents the standard reference value of the kth target indicator. The reference sequence includes the standard reference value of each target indicator. i (k) represents the standard data of the kth target indicator. The standard data matrix includes the standard data of each target indicator. ρ represents the resolution coefficient.

[0089] The substation carbon emission early warning indicator model obtained according to the comprehensive weight and the decision matrix includes:

[0090] D=ω ** E=(d1,d2,d3,……,d m );

[0091] Where D represents the substation carbon emission early warning indicator model, ω ** represents the comprehensive weight, E represents the decision matrix, d m Represents the gray comprehensive early warning index of each target indicator.

[0092] From the above description, it can be seen that through grey correlation analysis, the connections and laws between target indicators can be effectively identified under conditions of insufficient or even inaccurate data, and then the causal relationship between target indicators can be established, thereby improving the accuracy and reliability of early warning.

[0093] Please refer to Figure 4 Another embodiment of the present invention provides an early warning terminal for carbon emissions from a substation, comprising a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, each step in the above-mentioned early warning method for carbon emissions from a substation is implemented.

[0094] As can be seen from the above description, the beneficial effects of the present invention are as follows: Target indicators influencing substation carbon emissions are determined based on the PSR theoretical model. Experts, based on their professional knowledge and relevant experience, assess the degree of mutual influence between different target indicators, thereby obtaining subjective evaluation opinions. These subjective evaluation opinions, expressed in text form, are then quantified into numerical subjective weights using an evaluation language set. This approach reflects the inherent connections between target indicators in a complex system and provides a more comprehensive assessment of the importance of each target indicator within the indicator system. Furthermore, objective weights for target indicators are determined based on the entropy weight method, and a combined weighting is performed based on the subjective and objective weights to determine the overall weight for each target indicator. This combines the subjective opinions of experts with the objective results of scientific analysis, improving the objectivity and reliability of indicator weighting. Furthermore, considering the uncertainty and incompleteness of target indicator data, gray correlation analysis is introduced based on the degree of correlation between different target indicators to determine the decision matrix between the indicators, effectively helping to fill the data gaps for the target indicators. Furthermore, the early warning interval is determined using the Raida criterion, which avoids classification errors caused by subjective judgment within the early warning area and effectively improves the accuracy of the substation carbon emission early warning indicator model.

[0095] The embodiments of the present invention provide a substation carbon emissions early warning method and terminal, which can be applied to substation carbon emissions early warning scenarios. They can combine the subjective opinions of experts with the objective results of scientific analysis to improve the objectivity and reliability of indicator weighting. Furthermore, they can determine the connections and patterns between multiple indicators based on abnormal data to improve the accuracy of early warnings. The following is an illustration of this method using a specific embodiment:

[0096] Please refer to Figures 1 to 3 , embodiment 1 of the present invention is:

[0097] A method for early warning of carbon emissions from a substation, comprising:

[0098] S1. Determine the target indicators that affect substation carbon emissions based on the PSR theoretical model.

[0099] In some embodiments, as Figure 2As shown in Figure 1, the target indicators affecting substation carbon emissions include pressure indicator C1, status indicator C2, and response indicator C3. Pressure indicator C1 is used to characterize the carbon emission pressure or potential carbon emission pressure faced by the substation. It reflects the factors and sources of pressure that influence carbon emissions generated by the transmission and transformation project. Pressure indicator C1 can be considered from multiple perspectives. First, material consumption. The transportation phase involves the transportation of large-scale substation equipment, building materials, and construction equipment and facilities from production or storage sites to the construction site. This process consumes a large amount of energy and generates a large amount of carbon dioxide. Second, construction. From the start to completion of the construction phase, this series of processes generates a large amount of carbon emissions. These emissions primarily come from the energy consumption of construction machinery, the water used for construction, and the water, electricity, and gas used by on-site personnel for living and office work. For example, pressure indicator C1 may include material consumption C11, labor consumption C12, machinery consumption C13, and energy consumption C14. Status indicator C2 assesses the impact of substation equipment and operating conditions on carbon emissions. It focuses on basic transmission and transformation project operating parameters, equipment status, and operational quality, enabling effective maintenance and management strategies to reduce carbon emissions. These indicators primarily focus on carbon emissions generated during the daily operations of the transmission, substation, lighting, and HVAC systems. For example, status indicator C2 may include main transformer power loss C21, substation load C22, equipment aging level C23, and SF6 leakage rate C24. Response indicator C3 assesses the transmission and transformation project's response to and actions taken to address carbon emissions. It primarily includes emission reduction activities implemented, low-carbon technology application, and renewable energy utilization. It is divided into carbon sequestration and other carbon reduction measures. For example, response indicator C3 may include clean energy penetration C31, greening rate C32, and low-carbon technology application rate C33. A substation carbon emissions early warning indicator system is constructed based on pressure indicator C1, status indicator C2, and response indicator C3.

[0100] S2. Obtain subjective evaluation opinions of different experts on the degree of influence between the target indicators.

[0101] S3. Convert the subjective evaluation opinions into subjective weights using a preset evaluation language set.

[0102] Specifically, step S3 includes:

[0103] S31. Convert the subjective evaluation opinions into intuitive fuzzy numbers using a preset evaluation language set.

[0104] In some embodiments, a five-scale language set is used, based on the traditional real number scale, to describe the degree of influence between target indicators. The linguistic variables in the language set are: very high influence, high influence, medium influence, low influence, and no influence. To further quantify the linguistic information, corresponding conversion relationships are constructed between the linguistic variables and the intuitionistic fuzzy numbers (IFNs), resulting in the evaluation language set, as shown in Table 1.

[0105] Table 1 Evaluation language set

[0106]

[0107]

[0108] In some embodiments, based on the aforementioned evaluation language set, the subjective evaluation opinions obtained in step S2 should include the aforementioned language variables. For example, when evaluating the impact of target indicator 1 on target indicator 2, if expert 1's subjective evaluation opinion is very high, then its corresponding intuitive fuzzy number (IFN) is (0.9444, 0.0556, 0.00), and expert 2's subjective evaluation opinion is high, then its corresponding intuitive fuzzy number (IFN) is (0.803, 0.154, 0.00), and so on, until the corresponding intuitive fuzzy numbers (IFN) of all experts are obtained. Thus, the evaluation is based on the logic of human language expression, effectively ensuring the accuracy of the quantification of subjective opinions.

[0109] S32. Calculate the comprehensive impact value of each target indicator according to the intuitionistic fuzzy number.

[0110] The comprehensive impact value includes a first impact value that affects other target indicators and a second impact value that is affected by other target indicators.

[0111] Specifically, step S32 includes:

[0112] S321. Aggregate the intuitionistic fuzzy numbers of different experts for the same target indicator to obtain an aggregated intuitionistic fuzzy number:

[0113]

[0114] Among them, f ij represents the aggregated intuitionistic fuzzy number, represents the intuitionistic fuzzy number of the kth expert on the target indicator, represents the expected value of the intuitionistic fuzzy number, Indicates the fuzziness degree of the intuitionistic fuzzy number, represents the confidence of the intuitionistic fuzzy number, L represents the number of experts, and λ krepresents the aggregation weight of the kth expert; ij represents the relationship between the i-th target indicator and the j-th target indicator. That is, the intuitionistic fuzzy number weighted geometric average (IFWGA) operator is selected to achieve information aggregation.

[0115] S322. Obtain a direct influence matrix according to the aggregated intuitionistic fuzzy number, and defuzzify and standardize the direct influence matrix to obtain a standard influence matrix.

[0116] In some embodiments, after the aggregated intuitionistic fuzzy numbers IFN of all target indicators are calculated through step 31 above, a direct influence matrix is ​​formed using the aggregated intuitionistic fuzzy numbers IFN of the target indicators as matrix elements as follows:

[0117]

[0118] in, represents the direct impact matrix, Represents the aggregated intuitionistic fuzzy number of the target indicator, that is, the direct influence matrix The matrix elements in are

[0119] In some embodiments, this will directly affect the matrix Defuzzification is specifically performed as follows:

[0120]

[0121]

[0122] Among them, X represents the direct influence matrix after defuzzification, x ij express The aggregated intuitionistic fuzzy number after defuzzification, that is, the matrix element in the direct influence matrix X after defuzzification is x ij .

[0123] In some embodiments, the defuzzified direct influence matrix X is standardized as follows:

[0124]

[0125]

[0126] Among them, Y represents the standard influence matrix, y ij Represents the matrix elements in the standard influence matrix, i.e. y ij Represents x ij Normalized aggregated intuitionistic fuzzy number.

[0127] S323. Determine a comprehensive impact matrix based on the standard impact matrix:

[0128] T=[tij ] n×n =Y(IY) -1 ;

[0129] Among them, T represents the comprehensive impact matrix, t ij represents the impact value of target indicator i on target indicator j (the impact value includes direct impact value and indirect impact value), and t ij is the matrix element in the comprehensive influence matrix T, Y represents the standard influence matrix, and I represents the identity matrix.

[0130] S324: Determine the first impact value and the second impact value according to the comprehensive impact matrix:

[0131]

[0132]

[0133] Among them, R i represents the first impact value of the j-th target indicator, C j Represents the second impact value of the j-th target indicator.

[0134] In some embodiments, the centrality and causality of the target indicator can be determined based on the first influence value and the second influence value, wherein the centrality M i =R i +C j , cause degree D i =R i -C j .

[0135] S33. Determine the subjective weight of each target indicator according to the comprehensive impact value.

[0136] Specifically, step S33 includes:

[0137] S331: Calculate the original weight of each target indicator according to the first impact value and the second impact value, and normalize the original weight to obtain a subjective weight:

[0138]

[0139]

[0140] Among them, w sj represents the subjective weight of the jth target indicator, w' sj represents the original weight of the j-th target indicator.

[0141] S4. Determine the objective weight of each target indicator by using the entropy weight method.

[0142] Specifically, step S4 includes:

[0143] S41. Standardize the sample data of each target indicator to obtain a standard factor value.

[0144] In some embodiments, the sample data of the positive indicator is normalized:

[0145]

[0146] Normalize the sample data of negative indicators:

[0147]

[0148] in, represents the standard factor value of the j-th target indicator, x j Represents the sample data of the jth target indicator, x jmin Indicates the minimum value of the jth target indicator in several sample data, x jmax Represents the maximum value of the jth target indicator in several sample data.

[0149] S42. Calculate the proportion of the target indicator in the sample data according to the standard factor value.

[0150] In some embodiments, the proportion of the j-th target indicator in the i-th sample data can be calculated based on the standard factor value as follows:

[0151]

[0152] Among them, p ij Indicates the proportion of the j-th target indicator in the i-th sample data, It represents the dimensionless value of the i-th sample data of the j-th target indicator after normalization, i,j = 1, 2, ..., m, m represents the number of sample data, and n represents the number of target indicators in each sample data.

[0153] S43. Calculate the entropy value of each target indicator according to the proportion.

[0154] In some embodiments, step S43 is specifically as follows:

[0155]

[0156] Among them, E j Represents the entropy value of the j-th target indicator.

[0157] S44. Determine the objective weight of each target indicator according to the entropy value.

[0158] In some embodiments, step S44 is specifically as follows:

[0159]

[0160] Among them, w oj Represents the objective weight of the j-th target indicator, k=n.

[0161] S5. Perform combined weighting according to the subjective weight and the objective weight to obtain a comprehensive weight among the target indicators.

[0162] Specifically, step S5 includes:

[0163] S51. Perform combined weighting according to the subjective weight and the objective weight to obtain a combined weight.

[0164] In some embodiments, it is assumed that for n target indicators ω n There are l weighting methods, then the weight vector w obtained by the kth weighting method is k for:

[0165] w k =(ω1k,ω2k,ω3k,…,ω n k);

[0166] Where k = 1, 2,…, l.

[0167] In order to combine the advantages of each weighting method, it is necessary to combine all weighting methods to obtain the combined weight vector W of all methods. c :

[0168] W c =θ1w1+θ2w2+θ3w3+…+θ l w l ;

[0169] In this embodiment, in order to combine the advantages of the subjective weighting method and the objective weighting method of the target indicators, step S51 includes:

[0170] W c =θ1W1+θ2W2;

[0171] Among them, W c Represents the combined weight of all target indicators, W1 represents the subjective weight of all target indicators, and W2 represents the objective weight of all target indicators; θ1 and θ2 represent the combination coefficients, and θ1 and θ2 satisfy the constraints

[0172] S52: Calculate the combined weights according to the principle of the sum of squared deviations, and construct an objective function with the maximum sum of squared deviations as the goal.

[0173] Specifically, step S52 includes:

[0174]

[0175] Among them, J(ω) represents the objective function, S i (W c ) indicates W c The sum of squared deviations between the i-th combined weight method and other methods, i = 1, 2, 3, ..., m, ω j 、 and W c The elements in b ij Indicates the sample data value of the target indicator.

[0176] S53: Solve the objective function, and determine the comprehensive weights between the objective indicators according to the solution result.

[0177] In some embodiments, step S53 is specifically as follows:

[0178] Let matrix B be:

[0179]

[0180] Then the objective function J(ω) can be further expressed as J(ω)=ω T Bω=θ T ω T Bωθ, based on the optimal problem of the sum of squares of m deviations, can be converted to: maxF(θ)=θ T ω T The solution of Bωθ is that θ satisfies the following constraints:

[0181] Let B1 = ω T Bω, then if and only if θ is equal to the largest eigenvalue λ of the matrix B1 max The corresponding unit eigenvector θ * When F(θ) reaches its maximum value, the unit eigenvector θ is obtained. * Then, according to the unit eigenvector θ * Calculate the comprehensive weight vector ω * is: * =ωθ * , where ω=[W1 W2],

[0182] The comprehensive weight vector ω * Perform normalization to obtain the comprehensive weight

[0183] S6. Perform grey correlation analysis on the target indicators to obtain a decision matrix, and obtain a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix.

[0184] S7. Determine the warning interval of the warning value output by the substation carbon emission warning indicator model using the Raida criterion, and classify the warning interval into warning levels.

[0185] Please refer to Figure 3 , the second embodiment of the present invention is:

[0186] A method for early warning of carbon emissions from a substation is different from the first embodiment in that the specific steps of step S6 and step S7 are limited. Specifically, step S6 includes:

[0187] S61. Determine the original data matrix and the reference sequence according to the target indicator.

[0188] In some embodiments, if the current number of target indicators is 11, and there are sample data from 10 substations that require early warning analysis, the original data matrix is ​​a 10×11 matrix, and its matrix elements are the values ​​of the 11 target indicators in the 10 substations.

[0189] In some embodiments, the elements in the reference sequence are the optimal values ​​in different sample data of the same target indicator. The reference sequence is a sequence composed of the optimal values ​​and serves as a comparison standard for the same target indicator under different sample data.

[0190] S62. Perform dimensionless processing on the original data matrix to obtain a standard data matrix.

[0191] In some embodiments, assuming that there are m sample data and each sample data has n target indicators, the dimensionless processing is specifically as follows:

[0192]

[0193] Where i = 1, 2, ..., m, k = 1, 2, ..., n, x i (k) represents the original value of the kth target indicator in the i-th sample data, x' i (k) represents the dimensionless value of the kth target indicator in the i-th sample data. The standard data matrix is ​​composed of x' i (k) is composed of matrix elements.

[0194] S63. Calculate the grey relational coefficient of each target indicator according to the standard data matrix and the reference sequence.

[0195] Specifically, step S63 includes:

[0196]

[0197] Among them, ε i (k) represents the grey correlation coefficient of the kth target indicator, x' o (k) represents the standard reference value of the kth target indicator. The reference sequence includes the standard reference value of each target indicator. i (k) represents the standard data of the kth target indicator. The standard data matrix includes the standard data of each target indicator. ρ represents the resolution coefficient. In this embodiment, ρ=0.5.

[0198] S64. Determine a decision matrix according to the grey correlation coefficient.

[0199] In some embodiments, the decision matrix E is determined using the grey correlation coefficient obtained in step S63 as a matrix element as follows:

[0200]

[0201] S65: Obtain a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix. Specifically, step S65 includes:

[0202] D=ω ** E=(d1,d2,d3,……,d m );

[0203] Where D represents the substation carbon emission early warning indicator model, ω ** represents the comprehensive weight, E represents the decision matrix, d m Represents the gray comprehensive early warning index of each target indicator.

[0204] S7. Determine the warning interval of the warning value output by the substation carbon emission warning indicator model using the Raida criterion, and classify the warning interval into warning levels.

[0205] In some embodiments, the warning interval and warning level determined by the Raida criterion are shown in Table 2.

[0206] Table 2 Warning intervals and warning levels

[0207] Warning level Warning range Warning status Level 5 (-∞, μ-σ) Extremely high risk Level 4 (μ-σ, μ-0.5σ) High risk Level 3 (μ-0.5σ, μ+0.5σ) Medium risk Level 2 (μ+0.5σ, μ+σ) Low risk Level 1 (μ+σ,+∞) Risk-free

[0208] It should be noted that the Raida criterion (3σ criterion) is based on statistical principles and can more objectively determine that the warning value is within a reasonable threshold range for warning. Under the normal distribution, the range of its mean u plus or minus 3 times the standard deviation σ includes most (about 99.7%) of the data. Therefore, when applying the 3σ criterion to determine the threshold interval for carbon emission warnings, it can ensure that the vast majority of cases are covered, thereby taking into account the volatility and uncertainty of the data to a certain extent, and avoiding overly strict or overly loose threshold settings. Such threshold settings can more comprehensively reflect the actual situation and reduce the possibility of misjudgment. At the same time, it is also conducive to the formulation of more effective carbon emission management strategies and monitoring measures, providing a scientific basis for environmental protection and carbon emission reduction work.

[0209] Please refer to Figure 4 , the third embodiment of the present invention is:

[0210] A substation carbon emissions early warning terminal 300 includes a memory 301, a processor 302, and a computer program stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program, it implements the various steps of the substation carbon emissions early warning method of the above-mentioned embodiments 1 and 2.

[0211] In summary, the present invention provides a substation carbon emissions early warning method and terminal. This combines expert subjective assessments with objective weights determined using an entropy weighting method through an evaluation test method. This allows experts to participate in assessing the impact between different target indicators and comprehensively considers both direct and indirect impacts between target indicators, thereby more accurately determining the weights of each target indicator. This eliminates the potential bias and subjectivity inherent in subjective weighting, improves the objectivity and scientific nature of weight assignment, and mitigates potential limitations of objective weighting methods. Furthermore, after combining and weighting the target indicators, the gray correlation coefficients of the different target indicators are calculated based on the gray correlation analysis method's ability to handle incomplete and uncertain data. This characterizes the degree of correlation between the target indicators and generates a corresponding decision matrix. A substation carbon emissions early warning indicator model is determined based on the decision matrix and comprehensive weights, improving the accuracy and reliability of the model's early warnings. Finally, the 3σ principle is used to determine the early warning threshold range and warning level for substation carbon emissions, effectively avoiding classification errors caused by subjective judgment within the warning interval.

[0212] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's description and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for early warning of carbon emissions from a substation, characterized in that: include: Determine target indicators that affect substation carbon emissions based on the PSR theoretical model; Obtain subjective evaluation opinions of different experts on the degree of impact between the target indicators; Converting the subjective evaluation opinions into subjective weights through a preset evaluation language set; Determine the objective weight of each target indicator through entropy weight method; Performing combined weighting according to the subjective weight and the objective weight to obtain a comprehensive weight between the target indicators; Performing grey correlation analysis on the target indicators to obtain a decision matrix, and obtaining a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix; Determining a warning interval of a warning value output by the substation carbon emission warning indicator model by using the Raida criterion, and classifying the warning interval into warning levels; The converting of the subjective evaluation opinions into subjective weights using a preset evaluation language set includes: Converting the subjective evaluation opinions into intuitive fuzzy numbers through a preset evaluation language set; Calculating the comprehensive impact value of each target indicator according to the intuitionistic fuzzy number; Determining the subjective weight of each target indicator according to the comprehensive impact value; Determining the objective weight of each target indicator by the entropy weight method includes: Standardizing the sample data of each target indicator to obtain a standard factor value; Calculating the proportion of the target indicator in the sample data according to the standard factor value; Calculate the entropy value of each target indicator according to the proportion; Determining the objective weight of each target indicator according to the entropy value; The combined weighting according to the subjective weight and the objective weight to obtain the comprehensive weight between the target indicators includes: Performing combined weighting according to the subjective weight and the objective weight to obtain a combined weight; Calculating the combined weights according to the principle of the sum of squared deviations, and constructing an objective function with the maximum sum of squared deviations as the goal; Solving the objective function, and determining the comprehensive weights between the objective indicators according to the solution result; The step of performing grey correlation analysis on the target indicators to obtain a decision matrix, and obtaining a substation carbon emission early warning indicator model based on the comprehensive weight and the decision matrix includes: Determine the original data matrix and the reference sequence according to the target indicator; Performing dimensionless processing on the original data matrix to obtain a standard data matrix; Calculate the grey relational coefficient of each target indicator according to the standard data matrix and the reference sequence; Determine a decision matrix according to the grey correlation coefficient; A substation carbon emission early warning indicator model is obtained according to the comprehensive weight and the decision matrix.

2. The early warning method for carbon emissions from a substation according to claim 1, characterized in that: The comprehensive impact value includes a first impact value that affects other target indicators and a second impact value that is affected by other target indicators; Calculating the comprehensive impact value of each target indicator according to the intuitionistic fuzzy number includes: Aggregate the intuitionistic fuzzy numbers of different experts for the same target indicator to obtain the aggregated intuitionistic fuzzy number: ； in, represents the aggregated intuitionistic fuzzy number, represents the intuitionistic fuzzy number of the kth expert on the target indicator, , represents the expected value of the intuitionistic fuzzy number, Indicates the fuzziness of the intuitionistic fuzzy number, represents the confidence level of the intuitionistic fuzzy number, represents the aggregation weight of the k-th expert, k=1,2,…,L; Obtaining a direct influence matrix according to the aggregated intuitionistic fuzzy number, and defuzzifying and standardizing the direct influence matrix to obtain a standard influence matrix; Determine the comprehensive impact matrix based on the standard impact matrix: ； Where T represents the comprehensive impact matrix, t ij represents the impact value of target indicator i on target indicator j, Y represents the standard impact matrix, and I represents the unit matrix; Determine the first impact value and the second impact value according to the comprehensive impact matrix: ； ； in, R i represents the first impact value of the j-th target indicator, C j Represents the second impact value of the j-th target indicator.

3. The early warning method for carbon emissions from a substation according to claim 2, characterized in that: Determining the subjective weight of each target indicator according to the comprehensive impact value includes: The original weight of each target indicator is calculated according to the first impact value and the second impact value, and the original weight is normalized to obtain a subjective weight: ； ； in, w sj represents the subjective weight of the j-th target indicator, represents the original weight of the j-th target indicator.

4. The early warning method for carbon emissions from a substation according to claim 1, characterized in that: Calculating the grey relational coefficient of each target indicator according to the standard data matrix and the reference sequence includes: ； in, represents the grey relational coefficient of the kth target indicator, Represents the standard reference value of the kth target indicator. The reference sequence includes the standard reference value of each target indicator. Represents the standard data of the kth target indicator. The standard data matrix includes the standard data of each target indicator. represents the resolution coefficient; The substation carbon emission early warning indicator model obtained according to the comprehensive weight and the decision matrix includes: ； Where D represents the substation carbon emission early warning indicator model, represents the comprehensive weight, E represents the decision matrix, (d1, d2, d3, ..., d m ) represent the grey comprehensive early warning index of each target indicator.

5. A substation carbon emission early warning terminal, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, each step of the early warning method for carbon emissions from a substation as claimed in any one of claims 1 to 4 is implemented.

Images