An entropy-weighted TOPSIS energy potential evaluation method based on SCAD penalized regression

Abstract

The present invention relates to an entropy weight TOPSIS energy potential evaluation method based on SCAD penalty regression, comprising the following steps: step S1: determining a sample area, constructing an energy performance evaluation index system, and acquiring data; step S2: using the SCAD penalty regression method to establish the relationship between each evaluation index and energy consumption, and determining a new index system for energy performance evaluation; step S3: performing data standardization processing on the evaluation index in the new index system, and using the entropy value method to calculate the index weight under the new index system; step S4: using the TOPSIS model to calculate the energy performance evaluation index, and obtaining the final performance ranking. The method is conducive to the effective evaluation of energy potential.

Description

Technical Field

[0001] The invention relates to the field of energy potential assessment, and in particular to an entropy weight TOPSIS energy potential assessment method based on SCAD penalty regression. Background Art

[0002] Building an objective "clean, low-carbon, safe and efficient" energy system is an effective way to achieve efficient energy utilization. At present, in the evaluation of energy performance potential, subjective experience interferes too much in the process of system construction, and there is a lack of evaluation methods to effectively eliminate subjective interference. Summary of the invention

[0003] The purpose of the present invention is to provide an entropy weight TOPSIS energy potential evaluation method based on SCAD penalty regression, which is conducive to the effective evaluation of energy potential.

[0004] To achieve the above object, the technical solution adopted by the present invention is: an entropy weight TOPSIS energy potential evaluation method based on SCAD penalty regression, comprising the following steps:

[0005] Step S1: determine the sample area, construct the energy performance evaluation index system, and obtain data;

[0006] Step S2: Use the SCAD penalty regression method to establish the relationship between each evaluation index and energy consumption, and determine a new index system for energy performance evaluation;

[0007] Step S3: Perform data standardization on the evaluation indicators in the new indicator system, and use the entropy method to calculate the indicator weights under the new indicator system;

[0008] Step S4: Use the TOPSIS model to calculate the energy performance evaluation index and obtain the final performance ranking.

[0009] Furthermore, in step S1, n sample areas are determined, and an energy performance evaluation index system is constructed, wherein there are G evaluation indexes in total, and then data is acquired.

[0010] Furthermore, in step S2, the SCAD penalty regression method is used to establish the relationship between each evaluation index and energy consumption, and the specific method is:

[0011] For the linear regression equation:

[0012] y=Xβ+ε

[0013] Where y is energy consumption, X is each evaluation index, and β is the coefficient corresponding to each index; let z = X T y, Introducing the penalty function p j(·), then the equation containing the penalty term is:

[0014]

[0015] Among them, λ is the threshold parameter in the penalty function, G represents the number of indicators in the initial indicator system, β j represents the coefficient corresponding to the jth indicator;

[0016] Minimize the above equation; minimizing the above equation is equivalent to Minimize, then:

[0017] p λ (|θ|)=λ 2 -(|θ|-λ) 2 I(|θ|<λ)

[0018] Further we can get:

[0019] Among them, p λ (|·|) represents λp(|·|), which means that p(|·|) depends on λ; θ represents the parameter to be estimated in the equivalent formula, is the estimator of θ; I(·) represents the indicator function;

[0020] At the same time, the following equation is used for the continuously differentiable penalty function:

[0021]

[0022] The above equation is solved as:

[0023]

[0024] For the value of a, the solution is obtained by cross-validation method;

[0025] By introducing penalty terms, the coefficient characteristics of different indicators are shrunk respectively, and the indicators that are 0 after shrinkage are eliminated, so as to obtain a new indicator system including m evaluation indicators.

[0026] Furthermore, in step S3, data standardization is performed on the evaluation indicators in the new indicator system, specifically:

[0027] X=(x ij ) z×m (i=1,2,3,...,z; j=1,2,3,...,m)

[0028]

[0029] Among them, x ijis the jth index value of the ith region, there are n regions and m indicators in total; x′ ij For x ij The normalized value.

[0030] Furthermore, in step S3, the entropy method is used to calculate the indicator weights under the new indicator system, specifically:

[0031]

[0032] in, The entropy of the jth index is H j , p ij is x′ ij The proportion of all standardized values ​​in the sum, k is the parameter for calculating the entropy value;

[0033] The weight of each indicator is: and

[0034] Furthermore, in step S4, the TOPSIS model is used to calculate the energy performance evaluation index and obtain the final performance ranking, and the specific method is as follows:

[0035] For the energy performance evaluation index system, calculate the weighted matrix and determine the optimal solution and the worst solution

[0036] R=(r ij ) n×m ,r ij =ω j ·x′ ij (i=1,2,3,...,n; j=1,2,3,...,m)

[0037]

[0038]

[0039] Among them, r ij For x ij The value after weighted standardization; R is the corresponding weighted standardization matrix;

[0040] The Euclidean distance between each evaluation unit and the optimal solution and the worst solution is:

[0041]

[0042]

[0043] Calculate the energy performance evaluation index C of the i-th area i : And sort the indexes to get the final performance ranking.

[0044] Compared with the prior art, the present invention has the following beneficial effects: in the process of energy potential assessment, the present invention effectively eliminates subjective interference in the evaluation system and calculates the energy performance level of the target area, which is more realistic. Through this method, an energy performance index that is closer to reality can be obtained, and its potential can be effectively evaluated, which can provide data support for the formulation of medium- and long-term energy policies. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The present invention is a flowchart of a method implementation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0047] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0048] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0049] like Figure 1 As shown, this embodiment provides an entropy weight TOPSIS energy potential evaluation method based on SCAD penalty regression, comprising the following steps:

[0050] Step S1: Determine the sample area, build an energy performance evaluation index system, and obtain data.

[0051] Step S2: Use the SCAD penalty regression method to establish the relationship between each evaluation index and energy consumption, and determine a new index system for energy performance evaluation. The specific method is:

[0052] For the linear regression equation:

[0053] y=Xβ+ε

[0054] Where y is energy consumption, X is each evaluation index, and β is the coefficient corresponding to each index; let z = X T y, Introducing the penalty function pj (·), then the equation containing the penalty term is:

[0055]

[0056] Among them, λ is the threshold parameter in the penalty function, G represents the number of indicators in the initial indicator system, β j Represents the coefficient corresponding to the j-th indicator.

[0057] Minimize the above equation; minimizing the above equation is equivalent to Minimize, then:

[0058] p λ (|θ|)=λ 2 -(|θ|-λ) 2 I(|θ|<λ)

[0059] Further we can get:

[0060] Among them, p λ (|·|) represents λp(|·|), which means that p(|·|) depends on λ; θ represents the parameter to be estimated in the equivalent formula, is the estimator of θ; I(·) represents the indicator function.

[0061] At the same time, the following equation is used for the continuously differentiable penalty function:

[0062]

[0063] The above equation is solved as:

[0064]

[0065] For the value of a, the solution is obtained through the cross-validation method.

[0066] By introducing penalty terms, the coefficient characteristics of different indicators are shrunk respectively, and the indicators that are 0 after shrinkage are eliminated, so as to obtain a new indicator system including m evaluation indicators.

[0067] Step S3: Perform data standardization on the evaluation indicators in the new indicator system, and use the entropy method to calculate the indicator weights under the new indicator system.

[0068] Among them, the evaluation indicators in the new indicator system are processed by data standardization, specifically:

[0069] X=(x ij ) z×m (i=1,2,3,...,z; j=1,2,3,...,m)

[0070]

[0071] Among them, x ij is the jth index value of the ith region, there are n regions and m indicators in total; x′ ij For x ij The normalized value.

[0072] The entropy method is used to calculate the indicator weights under the new indicator system, specifically:

[0073]

[0074] in, The entropy of the jth index is H j , p ij is x′ ij The ratio of all normalized values ​​to the sum of all normalized values, k is the parameter for calculating the entropy value.

[0075] The weight of each indicator is: and

[0076] Step S4: Use the TOPSIS model to calculate the energy performance evaluation index and obtain the final performance ranking. The specific method is:

[0077] For the energy performance evaluation index system, calculate the weighted matrix and determine the optimal solution and the worst solution

[0078] R=(r ij ) n×m ,r ij =ω j ·x i ' j (i=1,2,3,...,n; j=1,2,3,...,m)

[0079]

[0080]

[0081] Among them, r ij For x ij The value after weighted standardization; R is the corresponding weighted standardization matrix.

[0082] The Euclidean distance between each evaluation unit and the optimal solution and the worst solution is:

[0083]

[0084]

[0085] Calculate the energy performance evaluation index C of the i-th area i : And sort the indexes to get the final performance ranking.

[0086] In this embodiment, the sample areas are Fujian Province, Guangdong Province, Jiangsu Province, and Zhejiang Province. An energy performance evaluation index system is constructed and data is described. The energy performance evaluation index system is shown in Table 1, and the specific data are shown in Tables 2-1 to 2-3:

[0087] Table 1: Energy performance evaluation system

[0088]

[0089]

[0090] Table 2-1: Raw data

[0091]

[0092]

[0093]

[0094] Table 2-2: Raw data

[0095]

[0096]

[0097] Table 2-3: Raw data

[0098]

[0099]

[0100] Based on the SCAD penalty regression equation, the relationship between each indicator and energy consumption is examined. The specific results are shown in Table 3:

[0101] Table 3: Index coefficient results based on SCAD penalized regression

[0102]

[0103] Based on the results in Table 2, we can see that the indicators with a coefficient of 0 are eliminated and a new indicator system is summarized, as shown in Table 4:

[0104] Table 4: Adjusted energy performance evaluation system

[0105]

[0106]

[0107] The adjusted energy performance evaluation indicators are processed by data standardization, and the specific results are shown in Table 5-1 and Table 5-2; and the entropy method is used to calculate the indicator weights under the evaluation system, and the specific results are shown in Table 6-1 and Table 6-2:

[0108] Table 5-1: Standardization

[0109]

[0110]

[0111] Table 5-2: Standardization

[0112]

[0113]

[0114]

[0115] Table 6-1: Weight Description (%)

[0116]

[0117] Table 6-2: Weight Description (%)

[0118]

[0119]

[0120] The TOPSIS model is used to calculate the energy performance index (i.e., relative proximity) and obtain the final performance ranking. The specific results are shown in Table 7:

[0121] Table 7: Energy performance index and ranking

[0122]

[0123]

[0124]

[0125] The above results can be used to analyze comprehensive energy performance. Through analysis, it can be found that in recent years, the energy development rankings of various provinces have changed greatly. From the perspective of the average energy performance index, Fujian has the best development, followed by Guangdong, Jiangsu has a large room for improvement, and Zhejiang has the largest room for improvement.

[0126] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. An entropy weighted TOPSIS energy potential assessment method based on SCAD penalized regression, characterized in that: The following steps are involved: Step S1: determine the sample area, construct the energy performance evaluation index system, and obtain data; Step S2: Use the SCAD penalty regression method to establish the relationship between each evaluation index and energy consumption, and determine a new index system for energy performance evaluation; Step S3: Perform data standardization on the evaluation indicators in the new indicator system, and use the entropy method to calculate the indicator weights under the new indicator system; Step S4: Calculate the energy performance evaluation index using the TOPSIS model and obtain the final performance ranking; In step S2, the SCAD penalty regression method is used to establish the relationship between each evaluation index and energy consumption, and the specific method is as follows: For the linear regression equation: y=Xβ+ε Where y is energy consumption, X is each evaluation index, and β is the coefficient corresponding to each index; let z = X T y, Introducing the penalty function p j (·), then the equation containing the penalty term is: Among them, λ is the threshold parameter in the penalty function, G represents the number of indicators in the initial indicator system, β j represents the coefficient corresponding to the jth indicator; Minimize the above equation; minimizing the above equation is equivalent to Minimize, then: p λ (θ)=λ 2 -(θ-λ) 2 I(θ<λ) Further we can get: Among them, p λ (·) represents λp(·), which means p(·) depends on λ; θ represents the parameter to be estimated in the equivalent formula, is the estimator of θ; I(·) represents the indicator function; At the same time, the following equation is used for the continuously differentiable penalty function: The above equation is solved as: For the value of a, the solution is obtained by cross-validation method; By introducing penalty terms, the coefficient characteristics of different indicators are shrunk respectively, and the indicators that are 0 after shrinkage are eliminated, so as to obtain a new indicator system including m evaluation indicators; In step S3, the entropy method is used to calculate the indicator weights under the new indicator system, specifically: in, The entropy of the jth index is H j , p ij For x i ' j The proportion of all standardized values ​​in the sum, k is the parameter for calculating the entropy value; The weight of each indicator is: and In step S4, the TOPSIS model is used to calculate the energy performance evaluation index and obtain the final performance ranking, and the specific method is as follows: For the energy performance evaluation index system, calculate the weighted matrix and determine the optimal solution and the worst solution R=(r ij ) n×m ,r ij =ω j ·x i ′ j (i=1,2,3,...,n;j=1,2,3,...,m) Among them, r ij For x ij The value after weighted standardization; R is the corresponding weighted standardization matrix; The Euclidean distance between each evaluation unit and the optimal solution and the worst solution is: Calculate the energy performance evaluation index C of the i-th area i : C i ∈[0,1], and sort the indexes to get the final performance ranking.

2. The entropy weighted TOPSIS energy potential assessment method based on SCAD penalized regression according to claim 1 is characterized in that: In the step S1, n sample areas are determined, and an energy performance evaluation index system is constructed, wherein there are G evaluation indexes in total, and then data is acquired.

3. The entropy weighted TOPSIS energy potential assessment method based on SCAD penalized regression according to claim 1 is characterized in that: In step S3, data standardization is performed on the evaluation indicators in the new indicator system, specifically: X=(x ij ) n×m (i=1,2,3,...,n;j=1,2,3,...,m) Among them, x ij is the jth index value of the ith region, there are n regions and m indicators in total; x i ' j For x ij The normalized value.

Images