Improved TOPSIS distributed power supply grid-connected power quality comprehensive evaluation method based on three-scale analytic hierarchy process

Abstract

The invention discloses an improved TOPSIS distributed power supply grid-connected power quality comprehensive evaluation method based on a three-scale analytic hierarchy process. The method comprises the following steps: step 1) obtaining a subjective weight by adopting the three-scale method to improve the analytic hierarchy process; (2) objective weights of all the indexes are solved through a CRITIC weighting method; 3) calculating a comprehensive weight based on a multiplication synthesis method; and 4) inputting the comprehensive weight value into the improved TOPSIS model to obtain a score sequence, and judging the quality of the electric energy. The analytic hierarchy process is improved through the three-scale method, the data set and calculation difficulty is reduced, and the evaluation timeliness is improved; a subjective and objective weight combination method is adopted, adverse effects caused by subjective weights and objective weights can be effectively weakened, TOPSIS is improved by using a grey relational degree method, and the problem that the traditional TOPSIS method cannot distinguish the advantages and disadvantages of indexes when the indexes are located on perpendicular bisectors of positive and negative ideal solutions is effectively solved.

Description

Technical Field

[0001] The invention relates to a comprehensive power quality evaluation method, in particular to an improved TOPSIS distributed power supply grid-connected power quality comprehensive evaluation method based on a three-scale analytic hierarchy process, and belongs to the technical field of comprehensive power quality evaluation. Background Art

[0002] The problem of energy consumption in the world is becoming increasingly prominent. While energy consumption has been alleviated, a series of problems have also arisen in the connection of new energy to the grid. Due to the volatility and intermittency of renewable energy, serious harmonic pollution and voltage fluctuations occur in the grid-connected area. The management of power quality requires consideration of many factors, and power quality assessment is a prerequisite for the management of power management issues. In this context, it is of great significance to propose an efficient, concise and accurate evaluation method to measure the quality of power.

[0003] Due to the intermittent and uncertain nature of the grid connection of a large number of distributed power sources in modern power systems; the transmission lines, transformers, fuses and other nonlinear devices have additional power loss and harmonic pollution during the transmission process; the randomness and complexity of the load access on the load side objectively exist, resulting in power quality problems that are different from traditional power quality problems. For example: the grid connection of distributed power sources mainly affects the harmonics and voltage fluctuations of the power grid, and has a relatively weak impact on three-phase imbalance. Most of the power quality assessment methods in the existing technology are also aimed at the transmission and power consumption stages, and there are fewer power quality assessments in the power generation stage. Therefore, a scientific and comprehensive assessment of the power quality of the grid-connected area of ​​distributed power sources has important engineering practice application significance. At this stage, the mainstream assessment method is still to comprehensively weight the subjective and objective weights, and use the comprehensive evaluation method to establish an assessment system to achieve the purpose of accurate assessment. However, due to the different comprehensive weighting methods, the comprehensive weight results generated are different, and the accuracy of the comprehensive assessment of power quality cannot meet the actual needs of the project.

[0004] Therefore, it is necessary to improve the accuracy and efficiency of the evaluation method by improving the means of weighting. The subjective weight determination methods in the prior art generally include the analytic hierarchy process and the priority diagram method. These two methods only consider the data size information, and do not consider the volatility and correlation of the data. Although the analytic hierarchy process can condense information and reduce the difficulty of data collection, it has a certain impact on the objective authenticity of the evaluation because it relies on expert judgment, so it is not suitable for use alone. The objective weight determination methods generally include the entropy method and the CRITIC (Criteria Importance Though IntercrieriaCorrelation) method. The entropy method only considers the differences between data, and does not consider the volatility and correlation of data, and contains too little information. The CRITIC method considers both the volatility of data and the correlation between data, and the weight is more objective. CRITIC contains enough information, but it requires a large data sample and actual problem domain. Therefore, the multiplication synthesis method is used to combine the subjective and objective weights, which can not only ensure the objectivity of the weights, but also reduce the impact of insufficient data samples.

[0005] A comprehensive evaluation needs to be conducted on the basis of subjective and objective weights. The comprehensive evaluation method usually adopts TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution distance method). TOPSIS can effectively utilize the original data information, but there is an inherent defect. When the index values ​​of the two evaluation objects are symmetrical about the line connecting the optimal solution and the worst solution, accurate results cannot be obtained. The grey correlation method considers the similarity of the development trends between factors, which can effectively solve the inherent defects of the TOPSIS method, thereby improving the TOPSIS method. Based on the analysis of the above background technology, the present invention considers using the three-scale method to improve the hierarchical analysis method, and improving TOPSIS based on the grey correlation method, and develops an improved TOPSIS distributed power supply grid-connected power quality comprehensive evaluation method based on the three-scale hierarchical analysis method. Summary of the invention

[0006] The purpose of the present invention is to provide an improved TOPSIS distributed power grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process. The existing analytic hierarchy process often fails to meet the actual needs of the project when constructing the judgment matrix, and needs to be adjusted many times. In order to improve the consistency of the judgment matrix, the three-scale method is introduced to improve the traditional hierarchy. Secondly, the correlation between the data is considered, and the CRITIC method is used to objectively assign weights. At the same time, a comprehensive weighting method is used to obtain the comprehensive weight, and the correlation between the subjective and objective weightings is improved. Finally, TOPSIS is improved based on the grey correlation method, which effectively solves the problem that the traditional TOPSIS method cannot distinguish the pros and cons of indicators when the indicators are on the median of the positive and negative ideal solutions.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] An improved TOPSIS distributed generation grid-connected power quality comprehensive evaluation method based on three-scale analytic hierarchy process, the steps are as follows:

[0009] Step 1) Use the three-scale method to improve the analytic hierarchy process to obtain the subjective weight α j ;

[0010] Step 2) Use the CRITIC weighting method to obtain the objective weight of each indicator;

[0011] Step 3) Calculate the comprehensive weight based on the multiplication synthesis method;

[0012] Step 4) Input the comprehensive weight value into the improved TOPSIS model to obtain the score ranking and judge the quality of power quality.

[0013] The purpose of the present invention can also be further achieved by the following technical measures:

[0014] In the above-mentioned improved TOPSIS distributed power generation grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process, the method for calculating the subjective weight in step 1) is:

[0015] (1) Introducing the optimal matrix: Using the three-scaling method (-1 0 1) to construct the judgment matrix A = (a ij ) n×n , where the matrix A=(a ij ) n×n elementa ij The meaning is as follows:

[0016]

[0017] According to the opinions of the expert database, the importance of each evaluation index is obtained, and the judgment matrix A is obtained. ij ) n×n , A is a real matrix with zeros on the main diagonal. If aij =-a ji , the matrix A is an antisymmetric matrix;

[0018] If the antisymmetric matrix A elements satisfy a ij =a im +a mj , m represents any row or column, then the matrix A is transitive; when the transitive matrix A and the matrix S satisfy the following formula:

[0019]

[0020] Where S = (b ij ) n×n , at this time, the S matrix is ​​called the optimal matrix of A;

[0021] (2) Introducing the consistency matrix, when S is the optimal matrix of A and satisfies A * =e s , e is a natural base, then the matrix A * is the consistency matrix of A, the matrix A * The following form:

[0022]

[0023] Find the matrix A * The largest eigenvector of :

[0024] A * φ j =λ max φ j ;

[0025] where v j A * The largest eigenvector of , λmax is the largest eigenroot; j Normalize it and get the subjective weight α j :

[0026]

[0027] In the above-mentioned improved TOPSIS distributed power grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process, the method for calculating the objective weight in step 2) is:

[0028] Let C j represents the amount of information contained in the jth indicator, C j The calculation formula is:

[0029]

[0030] In the formula, δ i C j The standard deviation, rij is the correlation coefficient, n is the number of evaluations of the same indicator; Where i = 1, 2, ... n, j = 1, 2, ... n, is the vector modulus, and the standard modulus error is

[0031] Obtain objective weight

[0032] In the above-mentioned improved TOPSIS distributed power grid-connected power quality comprehensive evaluation method based on three-scale analytic hierarchy process, the step 3) calculates the comprehensive weight w j The method is:

[0033]

[0034] The above-mentioned improved TOPSIS distributed power grid-connected power quality comprehensive evaluation method based on the three-scale hierarchical analysis method, the specific method of step 4) is:

[0035] Let the evaluation object be M = (M1, M2, ... M n ), the index set is D=(D1,D2,…D p ), the corresponding indicator value of the evaluation object is x ij (i=1,2,…,n;j=1,2,…,p), then the sample matrix x ij It is expressed as:

[0036]

[0037] According to the standardized formula of benefit index and the standardized formula of cost index, the indexes in the sample matrix are standardized to obtain the dimensionless matrix V;

[0038] Among them, the standardized formula of the efficiency index is:

[0039]

[0040] The standardized formula for the cost indicator is:

[0041]

[0042] In the formula, v ij Represents the standardized value of the indicator, max(x j ) and min(x j ) represents the maximum and minimum values ​​of the jth index;

[0043] The comprehensive weight w j Multiply it with the dimensionless matrix V to get the weighted decision matrix T;

[0044] Where V and T are:

[0045] V=(v ij ) n×n

[0046] T=(t ij ) n×n

[0047] T=w j ×V

[0048] where t ij is the weighted decision coefficient of each item;

[0049] According to the weighted decision matrix, the weighted Euclidean distance and grey correlation degree of each evaluation object to the positive and negative ideal solutions are solved;

[0050] According to the weighted decision matrix, the ideal solution S j + and negative ideal solution S j - Solution:

[0051]

[0052] The weighted Euclidean distance Sd from each evaluation object to the positive and negative ideal solutions i + and Sd i - for:

[0053]

[0054] Among them, w j is the comprehensive weight;

[0055] The i-th evaluation object M i With positive ideal solution S j + and negative ideal solution S j - The grey correlation coefficient e of the jth index ij + 、e ij - for:

[0056]

[0057] Where ξ is the resolution coefficient, 0≤ξ≤1;

[0058] Then the i-th evaluation object M i With positive ideal solution S j + and negative ideal solution S j - The grey relational degree of the jth index is:

[0059]

[0060] In the formula, E i + 、E i - are the i-th evaluation object and the positive ideal solution S j + and negative ideal solution S j - Grey relational degree of the jth index;

[0061] According to the following formula, the weighted Euclidean distance and grey correlation degree of each evaluation object to the positive and negative ideal solutions are dimensionless:

[0062]

[0063] This formula represents the corresponding element divided by the largest element among all elements, where G represents the weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions. i + and Sd i - , Grey relational degree E i + and E i - Any of the following:

[0064] The weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions i + , Sd i - And grey relational degree E i + 、E i - Combine to form the comprehensive feature vector of the i-th evaluation object, and the optimal distance V of the comprehensive feature vector of the i-th evaluation object i + and the worst distance V i - for:

[0065] V i + =α×Sd i - +(1-α)×E i + ,i=1,2,…,n

[0066] V i - =β×Sd i + +(1-β)×Ei - ,i=1,2,…,n

[0067] Where α and β are both non-negative real numbers;

[0068] According to the optimal distance and the worst distance of the comprehensive feature vector of each evaluation object, the comprehensive score of each evaluation object is solved according to the following formula:

[0069] Because of the use of extremely small indicators, η i The larger the value, the worse the evaluation object is. i The smaller the value, the better the evaluation object.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] Most of the existing technologies for evaluating power quality are focused on the transmission and consumption stages, and there are relatively few evaluation methods for the power generation stage. In addition, there are relatively few evaluation methods proposed for power quality issues caused by the grid connection of distributed power sources.

[0072] The present invention proposes an improved TOPSIS distributed photovoltaic power quality comprehensive evaluation method based on the three-scale analytic hierarchy process, which solves the problem of power quality evaluation in the distributed power grid-connected power generation stage. The proposed evaluation method can accurately and efficiently evaluate the power quality in the distributed power grid-connected power generation stage, and treat power quality problems such as harmonic pollution and voltage sag caused by grid connection based on the evaluation results. The five evaluation indicators proposed in the present invention cover all power quality problems caused by the distributed power grid connection, so the power quality can be accurately evaluated. The three-scale method of the present invention improves the analytic hierarchy process, reduces the data set and calculation difficulty, and improves the timeliness of the evaluation; the method of combining subjective and objective weights can effectively weaken the adverse effects of subjective weights and objective weights respectively, and improves TOPSIS by using the gray correlation method, which effectively solves the problem that the traditional TOPSIS method cannot distinguish the advantages and disadvantages of indicators when the indicators are on the median of the positive and negative ideal solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 is a flow chart of the method of the present invention;

[0074] Figure 2 It is a power quality system diagram. DETAILED DESCRIPTION

[0075] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0076] The flow chart of the present invention is as follows Figure 1 As shown in Figure 1, the power quality system is as follows: Figure 2 shown.

[0077] Taking 5 nodes in a certain area: A, B, C, D, and E, the measured voltage deviation, voltage fluctuation, frequency deviation, three-phase voltage unbalance, and total harmonic distortion rate are used as evaluation indicators. The measured data are shown in the following table.

[0078] Metrics_Object A B C D E Voltage deviation % 3.212 6.680 4.350 5.530 4.220 Three-phase unbalance % 0.83 1.36 1.35 1.74 1.83 Total harmonic distortion / Hz 1.72 5.28 2.67 3.36 4.57 Voltage fluctuation % 1.33 1.53 1.95 1.37 1.58 Frequency deviation % 0.0922 0.1562 0.1180 0.1787 0.1892

[0079] Subjective weight confirmation:

[0080] The judgment matrix A obtained by combining the three-scaling method is:

[0081]

[0082] At this time, the matrix A satisfies a ij =-a ji , so the matrix A is an antisymmetric matrix.

[0083] If the antisymmetric matrix A elements satisfy a ij =a im +a mj , then the matrix A is transitive. When the transitive matrix A and the matrix S satisfy the following formula:

[0084]

[0085] Where S = (b ij ) n×n , at this time, the S matrix is ​​called the optimal matrix of A, and the consistency matrix is ​​introduced. The optimal matrix S is obtained as:

[0086]

[0087] When S is the optimal matrix of A and satisfies A * =e s , then the matrix A * is the consistency matrix of A. * The mathematical definition of can be written as follows:

[0088] According to the relation A * =e S The obtained consistency matrix A * for:

[0089]

[0090] Since the consistency matrix A * With consistency requirements, no need to verify again, find the matrix A * The largest eigenvector of :

[0091] A * φ j =λ max φ j;

[0092] where φ j A * The largest eigenvector of j Normalize it and get the subjective weight α j .

[0093]

[0094] We get: j =(0.1170 0.3885 0.1170 0.1170 0.2640)

[0095] Objective weight confirmation:

[0096] Let C j represents the amount of information contained in the jth indicator, C j The calculation formula is:

[0097] (n is the number of evaluations of the same indicator)

[0098] In the formula, δ i is the standard deviation, r ij is the correlation coefficient.

[0099] and Represents the magnitude of a vector, where i=1,2,…n, j=1,2,…n.

[0100] C j The standard deviation of

[0101]

[0102] The calculated objective weight is: β j =[0.1635 0.1763 0.3198 0.1728 0.1676]

[0103] Comprehensive weight confirmation:

[0104]

[0105] The comprehensive weight is calculated as: w j =[0.1009 0.3614 0.1974 0.1067 0.2335].

[0106] Let the evaluation object be M = (M1, M2, ... M n ), the index set is D=(D1,D2,…D p ), the corresponding indicator value of the evaluation object is x ij(i=1,2,…,n;j=1,2,…,p), then the sample matrix x ij It can be expressed as:

[0107]

[0108] According to the standardized formula of benefit index and the standardized formula of cost index, the indexes in the sample matrix are standardized to obtain the dimensionless matrix V.

[0109] Among them, the standardized formula of the efficiency index is:

[0110]

[0111] The standardized formula for the cost indicator is:

[0112]

[0113] In the formula, v ij Represents the standardized value of the indicator, max(x j ) and min(x j ) represents the maximum and minimum value of the j-th indicator.

[0114] The comprehensive weight of each power quality index is multiplied by the dimensionless matrix to obtain a weighted decision matrix;

[0115] The comprehensive weight w j Multiply it with the dimensionless matrix V to get the weighted decision matrix T. Where V and T are:

[0116] V=(v ij ) n×p

[0117] T=(t ij ) n×p

[0118] T=w j ×V

[0119] where t ij is the weighted decision coefficient of each item.

[0120] According to the weighted decision matrix, the weighted Euclidean distance and grey correlation degree of each evaluation object to the positive and negative ideal solutions are solved;

[0121] According to the weighted decision matrix, the ideal solution S j + and negative ideal solution S j - Solution:

[0122]

[0123] The weighted Euclidean distance Sd from each evaluation object to the positive and negative ideal solutions i + and Sd i - for:

[0124]

[0125] Among them, w j is the comprehensive weight.

[0126] The i-th evaluation object M i With positive ideal solution S j + and negative ideal solution S j - Grey correlation coefficient e about the jth index ij + 、e ij - for:

[0127]

[0128] Where ξ is the resolution coefficient, 0≤ξ≤1;

[0129] Then the i-th evaluation object M i With positive ideal solution S j + and negative ideal solution S j - The grey relational degree of the jth index is:

[0130]

[0131] In the formula, E i + 、E i - are the i-th evaluation object and the positive ideal solution S j + and negative ideal solution S j - Grey relational degree of the jth indicator.

[0132] The weighted Euclidean distance and grey correlation degree of each evaluation object to the positive and negative ideal solutions are combined to obtain the optimal distance and the worst distance of the comprehensive feature vector of each evaluation object;

[0133] According to the following formula, the weighted Euclidean distance and grey correlation degree of each evaluation object to the positive and negative ideal solutions are dimensionless:

[0134]

[0135] This formula represents the corresponding element divided by the largest element among all elements, where G represents the weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions. i + and Sd i - , Grey relational degree E i + and E i - Any of the following:

[0136] The weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions i + , Sd i - And grey relational degree E i + 、E i - Combine to form the comprehensive feature vector of the i-th evaluation object, and the optimal distance V of the comprehensive feature vector of the i-th evaluation object i + and the worst distance V i - for:

[0137] V i + =α×Sd i - +(1-α)×E i + ,i=1,2,…,n

[0138] V i - =β×Sd i + +(1-β)×E i - ,i=1,2,…,n

[0139] Where α and β are both non-negative real numbers.

[0140] According to the optimal distance and the worst distance of the comprehensive feature vector of each evaluation object, the comprehensive score of each evaluation object is solved;

[0141] The comprehensive score of each evaluation object is calculated according to the following formula:

[0142] Because of the use of extremely small indicators, η i The larger the value, the worse the evaluation object is. i The smaller the value, the better the evaluation object.

[0143] η i=[0.1356 0.2798 0.1565 0.1779 0.2501]

[0144] The ranking and scoring of the five nodes A, B, C, D, and E are shown in the following table. The technology proposed in the present invention uses the three-standard method to improve the hierarchical analysis method, improves TOPSIS based on the grey correlation method, and establishes a simple and efficient evaluation system.

[0145] Evaluation object A B C D E Score 0.1356 0.2798 0.1565 0.1779 0.2501 Sorting 1 5 2 3 4

[0146] In addition to the above embodiments, the present invention may also have other implementation modes. Any technical solutions formed by equivalent replacement or equivalent transformation shall fall within the protection scope required by the present invention.

Claims

1. An improved TOPSIS method for comprehensive evaluation of distributed power quality based on three-scale analytic hierarchy process, characterized by: Here are the steps: Step 1) Use the three-scale method to improve the analytic hierarchy process to obtain the subjective weight α j ; Step 2) Use the CRITIC weighting method to obtain the objective weight of each indicator; Step 3) Calculating the comprehensive weight based on the multiplication synthesis method; Step 4) Input the comprehensive weight value into the improved TOPSIS model to obtain the score ranking and judge the quality of power quality.

2. The improved TOPSIS distributed power supply grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process according to claim 1 is characterized in that: The method for calculating the subjective weight in step 1) is: (1) Introducing the optimal matrix: Using the three-scaling method (-101) to construct the judgment matrix A = (a ij ) n×n , where the matrix A=(a ij ) n×n element a ij The meaning is as follows: According to the opinions of the expert database, the importance of each evaluation index is obtained, and the judgment matrix A=(a ij ) n×n , A is a real matrix with 0 on the main diagonal, if a ij =-a ji , the matrix A is an antisymmetric matrix; If the antisymmetric matrix A elements satisfy a ij =a im +a mj , m represents any row or column, then the matrix A is transitive; when the transitive matrix A and the matrix S satisfy the following formula: Where S=(b ij ) n×n , at this time the S matrix is ​​called the optimal matrix of A; (2) Introducing the consistency matrix, when S is the optimal matrix of A and satisfies A * =e s , e is a natural base, then the matrix A * is the consistency matrix of A, and the matrix A* is as follows: Obtain matrix A * The largest eigenvector of : A * f j =λ max f j ; where φ j A * The largest eigenvector of ,λmax is the largest eigenroot; j Perform normalization to obtain the subjective weight α j :

3. The improved TOPSIS distributed power generation grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process according to claim 1 is characterized in that: The method for calculating the objective weight in step 2) is: Let C j represents the amount of information contained in the jth indicator, C j The calculation formula is: Where, δ i C j The standard deviation, r ij is the correlation coefficient, n is the number of evaluations of the same indicator; Where, i = 1, 2, ... n, j = 1, 2, ... n, is the vector modulus, and the standard modulus error is Obtain objective weight 4. The improved TOPSIS distributed power generation grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process as claimed in claim 1 is characterized in that: Step 3) calculates the comprehensive weight w j The method is:

5. The improved TOPSIS distributed power generation grid-connected power quality comprehensive evaluation method based on the three-scale analytic hierarchy process as claimed in claim 1 is characterized in that: The specific method of step 4) is: Let the evaluation object be M=(M1,M2,…M n ), the index set is D=(D1,D2,…D p ), the indicator value corresponding to the evaluation object is x ij (i=1,2,…,n;j=1,2,…,p), then the sample matrix x ij Expressed as: According to the standardized formula of benefit index and the standardized formula of cost index, the indicators in the sample matrix are standardized to obtain the dimensionless matrix V; Among them, the standardized formula of the efficiency index is: The standardized formula for cost indicators is: Where, v ij Represents the standardized value of the indicator, max(x j ) and min(x j ) represents the maximum and minimum values ​​of the j-th index; The comprehensive weight w j Multiply it with the dimensionless matrix V to get the weighted decision matrix T; Where V and T are: V=(v ij ) n×n T=(t ij ) n×n T=w j ×V where t ij is the weighted decision coefficient of each item; According to the weighted decision matrix, the weighted Euclidean distance and grey relational degree of each evaluation object to the positive and negative ideal solutions are solved; According to the weighted decision matrix, the ideal solution S j + and negative ideal solution S j - Solution: The weighted Euclidean distance Sd from each evaluation object to the positive and negative ideal solutions i + and Sd i - for: Among them, w j is the comprehensive weight; The i-th evaluation object M i With the positive ideal solution S j + and negative ideal solution S j - Regarding the grey relational coefficient e of the jth index ij + 、e ij - for: Where ξ is the resolution coefficient, 0≤ξ≤1; Then the i-th evaluation object M i With the positive ideal solution S j + and negative ideal solution S j - The grey relational degree of the jth indicator is: Where, E i + 、E i - are the i-th evaluation object and the positive ideal solution S j + and negative ideal solution S j - Grey relational degree of the jth indicator; According to the following formula, the weighted Euclidean distance and grey relational degree of each evaluation object to the positive and negative ideal solutions are dimensionless: Where G represents the weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions i + and Sd i - , Grey relational degree E i + and E i - Any of the following: The weighted Euclidean distance Sd from the i-th evaluation object to the positive and negative ideal solutions i + 、Sd i - and grey relational degree E i + 、E i - Combine to form the comprehensive feature vector of the i-th evaluation object, and the optimal distance V of the comprehensive feature vector of the i-th evaluation object i + and the worst distance V i - for: V i + =α×Sd i - +(1-a)×E i + ,i=1,2,…,n V i - =β×Sd i + +(1-β)×E i - ,i=1,2,…,n Where α and β are both non-negative real numbers; According to the optimal distance and the worst distance of the comprehensive feature vector of each evaluation object, the comprehensive score of each evaluation object is solved according to the following formula: Because of the use of extremely small indicators, η i The larger the value, the worse the evaluation object is. On the contrary, η i The smaller the value, the better the evaluation object.

Images