Enterprise carbon asset management comprehensive evaluation method based on entropy weight method and ANFIS

Abstract

The invention relates to the field of carbon asset management, in particular to an enterprise carbon asset management comprehensive evaluation method based on an entropy weight method and ANFIS, and the method comprises the steps: firstly obtaining a to-be-evaluated enterprise sample set, and carrying out the data preprocessing of the to-be-evaluated enterprise sample set, so as to obtain a target data set; the enterprise sample set comprises transactions, technologies, data, social responsibilities and personnel, calculating index weights and information entropies for the target data set to construct an initial evaluation system, constructing a fuzzy neural network model based on the initial evaluation system, and evaluating the enterprise sample set to be evaluated by using the fuzzy neural network model. According to the scheme, the technical problem that an existing evaluation index system cannot comprehensively evaluate carbon asset management can be solved.

Description

Technical Field

[0001] The present invention relates to the field of carbon asset management, and in particular to a comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS. Background Art

[0002] At present, multiple carbon emission trading systems have been formed around the world. my country has launched seven carbon trading pilot provinces since 2011. On July 16, 2021, the national carbon emission trading market was officially opened. China's carbon market is expected to become the world's largest carbon trading market. Low-carbon transformation will be a long and complex process, which requires enterprises to change from "high investment, high emissions" to "low investment, low emissions".

[0003] Carbon asset management has become a key link in corporate management, because carbon assets run through the entire process of carbon management, from carbon materials that directly cause carbon emissions to carbon quota trading to carbon budget funding forecasts. Lack of effective carbon asset management may directly affect the company's carbon trading strategy and indirectly bring about many problems such as compliance risks, financial risks, emission reduction technology risks and market competition risks. Therefore, effective and systematic carbon asset management evaluation methods are needed to help companies identify and optimize their carbon management strategies.

[0004] Based on this, there is an urgent need for a comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS to solve the problem that the existing evaluation index system is not comprehensive in assessing carbon asset management. Summary of the invention

[0005] In order to solve the technical problem that the existing evaluation index system is not comprehensive in evaluating carbon asset management, the embodiment of this specification provides a comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS.

[0006] The embodiment of this specification provides a comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS, including:

[0007] Acquire a sample set of enterprises to be evaluated, and perform data preprocessing on the sample set of enterprises to be evaluated to obtain a target data set;

[0008] Calculating indicator weights and information entropy based on the target data set to construct an initial evaluation system;

[0009] Constructing a fuzzy neural network model based on the initial evaluation system;

[0010] The fuzzy neural network model is used to evaluate the enterprise sample set to be evaluated.

[0011] In one possible design, the sample set of enterprises to be evaluated includes transactions, technology, data, social responsibility and personnel.

[0012] In a possible design, the data preprocessing of the enterprise sample set to be evaluated to obtain a target data set includes:

[0013] The negative indicators and the single-point optimal indicators in the enterprise sample set to be evaluated are processed respectively to obtain an intermediate data set;

[0014] The intermediate target data set is normalized to obtain the target data set.

[0015] In a possible design, the target data set is determined by the following formula:

[0016]

[0017] In a possible design, the calculating the indicator weight and information entropy based on the target data set to construct an initial evaluation system includes:

[0018] Performing matrix processing on the target data set to obtain a data matrix;

[0019] Based on the data matrix, the weight of each enterprise sample to be evaluated under each indicator is calculated respectively;

[0020] Based on the data matrix, the indicator weight of each enterprise sample to be evaluated under each indicator is calculated respectively;

[0021] A comprehensive score is calculated based on the indicator weights, and the initial evaluation system is constructed according to the comprehensive score.

[0022] In one possible design, the information entropy is determined by the following formula:

[0023]

[0024] In the formula, E j represents information entropy, r j represents each enterprise sample to be evaluated, p ij It represents the weight of the i-th sample under the j-th indicator, n represents the total number of enterprise samples to be evaluated, and m represents the total number of enterprise sample indicators to be evaluated.

[0025] In one possible design, the specific gravity is determined by the following formula:

[0026]

[0027] In a possible design, the indicator weight is determined by the following formula:

[0028]

[0029] In the formula, w j Indicates the weight of each indicator.

[0030] In one possible design, the comprehensive score is determined by the following formula:

[0031]

[0032] In the formula, S i Indicates the comprehensive score.

[0033] In a possible design, the fuzzy neural network model is an ANFIS model.

[0034] The embodiments of this specification provide a comprehensive evaluation method for enterprise carbon asset management based on the entropy weight method and ANFIS. First, a sample set of enterprises to be evaluated is obtained, and data preprocessing is performed on the sample set of enterprises to be evaluated to obtain a target data set. The enterprise sample set includes transactions, technology, data, social responsibility and personnel. Then, the indicator weights and information entropy of the target data set are calculated to construct an initial evaluation system. A fuzzy neural network model is constructed based on the initial evaluation system. The fuzzy neural network model is used to evaluate the sample set of enterprises to be evaluated. Therefore, the above scheme can solve the technical problem that the existing evaluation index system is not comprehensive in evaluating carbon asset management. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 It is a flow chart of a method for comprehensive evaluation of enterprise carbon asset management based on entropy weight method and ANFIS provided in an embodiment of this specification;

[0037] Figure 2 It is a model flow chart of a comprehensive evaluation of enterprise carbon asset management based on entropy weight method and ANFIS provided in an embodiment of this specification;

[0038] Figure 3 It is a structural diagram of ANFIS provided in an embodiment of this specification. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the embodiments of this specification clearer, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this specification.

[0040] Please refer to Figure 1 The embodiment of this specification provides a comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS, the method comprising:

[0041] Step 100: Obtain a sample set of enterprises to be evaluated, and perform data preprocessing on the sample set of enterprises to be evaluated to obtain a target data set;

[0042] Step 102: Calculate the indicator weight and information entropy based on the target data set to build an initial evaluation system;

[0043] Step 104: constructing a fuzzy neural network model based on the initial evaluation system;

[0044] Step 106: Evaluate the sample set of enterprises to be evaluated using the fuzzy neural network model

[0045] The embodiments of this specification provide a comprehensive evaluation method for enterprise carbon asset management based on the entropy weight method and ANFIS. First, a sample set of enterprises to be evaluated is obtained, and data preprocessing is performed on the sample set of enterprises to be evaluated to obtain a target data set. The enterprise sample set includes transactions, technology, data, social responsibility and personnel. Then, the indicator weights and information entropy of the target data set are calculated to construct an initial evaluation system. A fuzzy neural network model is constructed based on the initial evaluation system. The fuzzy neural network model is used to evaluate the sample set of enterprises to be evaluated. Therefore, the above scheme can solve the technical problem that the existing evaluation index system is not comprehensive in evaluating carbon asset management.

[0046] like Figure 2As shown, the embodiment of this specification provides a comprehensive evaluation (prediction) method based on the entropy weight method and adaptive neural fuzzy inference system (ANFIS) in five dimensions of comprehensive transactions, technology, data, social responsibility and personnel. This method takes into account the reality that the carbon asset management evaluation system contains many uncertainties and ambiguities. With the help of the advantages of ANFIS combined with the interpretability of fuzzy logic and the learning ability of neural networks, as well as the objectivity of the entropy weight method in determining weights, an adaptive, high-precision, highly interpretable, and comprehensive enterprise carbon asset management evaluation method is constructed. It can be quickly deployed and easily adjusted, does not require huge computing resources like deep learning, and provides a low-cost, lightweight solution.

[0047] In one embodiment of the present specification, the enterprise sample set to be evaluated includes transactions, technology, data, social responsibility, and personnel.

[0048] In some embodiments, the enterprise sample set includes transactions, technology, data, social responsibility, and personnel, and the data is shown in Table 1 (showing primary indicators, secondary indicators, and indicator definitions or formulas).

[0049] Table 1 Enterprise Carbon Asset Management Evaluation Index System

[0050]

[0051]

[0052] In one embodiment of the present specification, data preprocessing is performed on the enterprise sample set to be evaluated to obtain a target data set, including:

[0053] The negative indicators and single-point optimal indicators in the enterprise sample set to be evaluated are processed separately to obtain an intermediate data set;

[0054] The intermediate data set is normalized to obtain the target data set.

[0055] In some embodiments, the negative index is determined by the following formula:

[0056]

[0057] The single point optimal index is determined by the following formula:

[0058]

[0059] In the formula, α represents the optimal value of the indicator in the sample set of enterprises to be evaluated.

[0060] In this embodiment, α is the optimal value of the indicator in the enterprise sample set to be evaluated, and the farther away from this point, the smaller the indicator.

[0061] In one embodiment of the present specification, the target data set is determined by the following formula:

[0062]

[0063] In this embodiment, all the positively transformed data are normalized to convert indicators of different dimensions or magnitudes to the same scale in order to perform direct comparison and calculation in comprehensive evaluation and analysis.

[0064] In one embodiment of the present specification, the indicator weight and information entropy are calculated based on the target data set to construct an initial evaluation system, including:

[0065] Matrix the target data set to obtain a data matrix;

[0066] Based on the data matrix, the weight of each enterprise sample to be evaluated under each indicator is calculated separately;

[0067] Based on the data matrix, the indicator weights of each enterprise sample to be evaluated under each indicator are calculated separately;

[0068] The comprehensive score is calculated based on the indicator weights, and the initial evaluation system is constructed based on the comprehensive score.

[0069] In one embodiment of the present specification, information entropy is determined by the following formula:

[0070]

[0071] In the formula, E j represents information entropy, r j represents each enterprise sample to be evaluated, p ij It represents the weight of the i-th sample under the j-th indicator, n represents the total number of enterprise samples to be evaluated, and m represents the total number of enterprise sample indicators to be evaluated.

[0072] In one embodiment of the present specification, the specific gravity is determined by the following formula:

[0073]

[0074] In one embodiment of the present specification, the indicator weight is determined by the following formula:

[0075]

[0076] In the formula, w j Indicates the weight of each indicator.

[0077] In one embodiment of the present specification, the comprehensive score is determined by the following formula:

[0078]

[0079] In the formula, S i Indicates the comprehensive score.

[0080] In this embodiment, the smaller the entropy value, the greater the weight w j The larger the value is, the more important the indicator is in the evaluation system. According to the different degrees of membership, the evaluation output (based on the percentage comprehensive score) is divided into five levels: poor, bad, medium, good and ideal as output distinctions. The data is shown in Table 2 (showing the level, evaluation level and evaluation output).

[0081] Table 2 Enterprise Carbon Asset Management Evaluation Domain Set

[0082]

[0083] In one embodiment of the present specification, the fuzzy neural network model is an ANFIS model.

[0084] In this embodiment, ANFIS consists of five functional modules: input layer, membership layer, rule layer, decision layer and output layer. Figure 3 The figure shows a first-order Sugeno circuit with two inputs x and y and a single output.

[21] Type fuzzy system.

[0085] This combination maximizes the advantages of each. On the one hand, ANFIS can handle nonlinear relationships and reveal the complex correlations between indicators, making up for the shortcomings of the entropy weight method, such as insufficient sensitivity to data changes and difficulty in adapting to dynamic environments. At the same time, it can also handle uncertainty and ambiguity, providing more flexible evaluation results. On the other hand, the entropy weight method reduces ANFIS's dependence on the "experience" of training samples through the objectivity of calculating indicator weights through information entropy, avoiding excessive influence of subjective factors.

[0086] ANFIS overcomes the black box characteristics of simple neural networks and the incompleteness and roughness of reasoning rules in the fuzzy reasoning process, simplifies data processing to the greatest extent, and has the characteristics of self-adaptation, self-organization and self-learning.

[0087] Based on the definition of carbon assets, the composition of carbon asset management and practical considerations, this method constructs a carbon asset management evaluation index system that combines transactions, technology, data, social responsibility and personnel to ensure the multi-dimensionality, effectiveness and complexity of the evaluation. A model that combines the entropy weight method with an artificial intelligence algorithm is used to map the five first-level indicators and their sub-attributes with the evaluation results. By continuously training the model, high-precision evaluation simulation results are obtained to achieve its application purpose.

[0088] See also Figure 3, in some implementations, x and y are input variables, forming two if-then rules:

[0089] Rule 1: If x is A1 and y is B1, then f_1=p_1x+q_1y+r_1;

[0090] Rule 2: If x is A2 and y is B2, then f_2=p_2x+q_2y+r_2;

[0091] Among them, p i ,q i 、r i is a variable coefficient, the lines between nodes only represent the signal flow direction, the square nodes represent nodes with adjustable parameters, and the round nodes represent nodes without adjustable parameters.

[0092] The various layers of the ANFIS model are described as follows:

[0093] Layer 1: Fuzzify the input variables and output the membership of the corresponding fuzzy set:

[0094]

[0095] Where: x is the input of node i; μ Ai (x) is a membership function, which is usually a bell curve, cosine function or Gaussian function. This method uses Gaussian function to express as

[0096]

[0097] In the formula, c i and σ i are all parameters set by the Gaussian function. Similarly, μ Bi (y) is the membership function of the associated variable Bi.

[0098] Layer 2: Implement fuzzy set operations for the conditional part, expressed as π, and output the applicability of each rule:

[0099]

[0100] Layer 3: Classification processing is performed to normalize the applicability of each rule, which is the usage frequency corresponding to the rule, used to indicate the triggering probability of the rule. For example, the normalized applicability of the i-th rule is:

[0101]

[0102] In the formula, is the normalized trigger strength.

[0103] Layer 4: Defuzzification operation, calculating the output of each rule:

[0104]

[0105] Layer 5: Fixed nodes, obtain accurate output through defuzzification, that is, weighted average of the results of each rule, and the final output is:

[0106]

[0107] In some embodiments, all secondary indicators of the training samples are used as input items, and the comprehensive scores are used as output items, and the fuzzy neural recognition network is trained using MATLAB software;

[0108] Substitute the test samples to be identified into the trained ANFIS comprehensive evaluation model for testing (prediction), and determine the carbon asset management level of the enterprise to be identified based on the system output value. Analyze and discuss the evaluation results.

[0109] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this specification, rather than to limit them. Although this specification has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of this specification.

Claims

1. A comprehensive evaluation method for enterprise carbon asset management based on entropy weight method and ANFIS, characterized in that: include: Acquire a sample set of enterprises to be evaluated, and perform data preprocessing on the sample set of enterprises to be evaluated to obtain a target data set; Calculating indicator weights and information entropy based on the target data set to construct an initial evaluation system; Constructing a fuzzy neural network model based on the initial evaluation system; The fuzzy neural network model is used to evaluate the enterprise sample set to be evaluated.

2. The method according to claim 1, characterized in that The sample set of enterprises to be evaluated includes transactions, technology, data, social responsibility and personnel.

3. The method according to claim 2, characterized in that The data preprocessing of the enterprise sample set to be evaluated to obtain a target data set includes: The negative indicators and the single-point optimal indicators in the enterprise sample set to be evaluated are processed respectively to obtain an intermediate data set; The intermediate data set is normalized to obtain the target data set.

4. The method according to claim 3, characterized in that The target data set is determined by the following formula:

5. The method according to claim 4, characterized in that The calculating the indicator weight and information entropy based on the target data set to construct an initial evaluation system includes: Performing matrix processing on the target data set to obtain a data matrix; Based on the data matrix, the weight of each enterprise sample to be evaluated under each indicator is calculated respectively; Based on the data matrix, the indicator weight of each enterprise sample to be evaluated under each indicator is calculated respectively; A comprehensive score is calculated based on the indicator weights, and the initial evaluation system is constructed according to the comprehensive score.

6. The method according to claim 5, characterized in that The information entropy is determined by the following formula: In the formula, E j represents information entropy, r j represents each enterprise sample to be evaluated, p ij It represents the weight of the i-th sample under the j-th indicator, n represents the total number of enterprise samples to be evaluated, and m represents the total number of enterprise sample indicators to be evaluated.

7. The method according to claim 6, characterized in that The specific gravity is determined by the following formula:

8. The method according to claim 7, characterized in that The indicator weight is determined by the following formula: In the formula, w j Indicates the weight of each indicator.

9. The method according to claim 8, characterized in that The composite score is determined by the following formula: In the formula, S i Indicates the comprehensive score.

10. The method according to claim 9, characterized in that The fuzzy neural network model is an ANFIS model.

Images