Intelligent grid comprehensive evaluating system based on analytic hierarchy process and fuzzy comprehensive evaluation

[0021] The present invention is a smart grid comprehensive evaluation system based on analytic hierarchy process and fuzzy comprehensive evaluation. It is composed of a smart grid evaluation system architecture, smart grid key evaluation indicators and a smart grid evaluation method. The specific implementation steps are as follows:

[0022] (1) Comprehensively sort out the overall framework of smart grid construction, grasp the profound connotation and development direction of smart grid construction, and establish a comprehensive smart grid comprehensive evaluation system architecture;

[0023] (2) Collect and sort out the basic data related to the smart grid, process the basic data related to the smart grid evaluation, and determine the key evaluation indicators of the smart grid;

[0024] (3) Based on the analytic hierarchy process, the basic weights of the evaluation indicators at each level of the smart grid are determined, and on this basis, the Delphi method is used to correct the basic weights to obtain the final weight vector;

[0025] (4) Using the fuzzy comprehensive evaluation method and according to the comprehensive opinions, realize the comprehensive evaluation of the development level of smart grid in different regions.

[0026] The structure of the smart grid evaluation system described in the present invention adopts the analytic hierarchy process, and according to the connection and difference between various smart grid characteristics, these characteristics are classified and layered to form a complete system including multiple levels. The key evaluation indicators of the smart grid can reflect the profound connotation of the smart grid from different angles, can fully reflect the development level of the smart grid, and have different index weights. The different index weights mentioned above are determined by using the AHP to determine the basic weights, and then corrected by the Delphi method to obtain the final weights. The key evaluation indicators of the smart grid include 3 first-level evaluation indicators, 19 second-level indicators and 88 third-level indicators. The smart grid evaluation method mainly includes: analytic hierarchy process, Delphi method and fuzzy comprehensive evaluation method. The fuzzy comprehensive evaluation method uses the opinions given by experts as the overall evaluation matrix, and the evaluation results are fully combined with the comprehensive opinions given by experts.

[0027] The establishment process of the present invention is as follows figure 1 As shown, the specific implementation steps are as follows:

[0028] (1) Establishment of the smart grid comprehensive evaluation system architecture.

[0029] like figure 2 As shown, the definition of my country's strong smart grid: "Smart grid refers to the UHV grid as the backbone grid, supported by the communication information platform, with the characteristics of informatization, automation, and interaction, including power generation, transmission, transformation, and distribution. , power consumption and dispatching, covering all voltage levels, and realizing a modern power grid with a high degree of integration of "power flow, information flow, and business flow."

[0030] Carefully sort out the strong smart grid technology framework system, fully grasp the future development trend of smart grid construction, and conclude: "Smart grid infrastructure construction situation", "Smart grid support technology research situation" and "Direct and indirect effects of smart grid construction " is the key point to reflect the development level of smart grid, thus initially establishing the basic framework of smart grid evaluation consisting of three categories of indicators: basic, technical and effect, that is, the first-level evaluation indicators of the evaluation system are: A Technical indicators and C effect indicators. The content of smart grid construction mainly includes six major links of the power system and a platform. Based on the analytic hierarchy process, the A basic index is determined to include: A1 power generation, A2 power transmission, A3 power transformation, A4 power distribution, A5 power consumption, A6 dispatching and A7 communication information There are 7 secondary indicators in total, and the determination of B technical indicators includes: B1 power generation, B2 power transmission, B3 power transformation, B4 power distribution, B5 power consumption, B6 dispatching and B7 communication information, a total of 7 secondary indicators. Comprehensively understand the profound connotation basis of strong smart grid, study the influence and effect of smart grid construction, and determine the five major connotations of smart grid as the secondary indicators under the C effect index based on the analytic hierarchy process, that is, the C effect index includes: C1Strong and reliable , C2 clean and environmentally friendly, C3 transparent and open, C4 friendly interaction, C5 economical and efficient five major categories of secondary indicators. Thus, the evaluation system architecture of the smart grid is preliminarily determined.

[0031] (2) Determine the key evaluation indicators of the smart grid.

[0032] To conduct smart grid assessment, the premise is to determine the key technical indicators for evaluating smart grid. Based on the initially established smart grid evaluation system architecture, all the second-level indicators that have been determined are refined, and a total of 88 third-level indicators under the 19 second-level indicators of the evaluation system are determined in line with the principles of science, rationality, objectivity and comprehensiveness.

[0033]Among them, A basic index includes 31 third-level indicators, B technical index includes 28 third-level indicators, and C effect index includes 29 third-level indicators. like figure 2 shown.

[0034] (3) Determine the basic weight of smart grid evaluation indicators.

[0035] Comparison matrix build:

[0036] Using the Analytic Hierarchy Process (AHP), select the indicators under the same level of the same indicator set (such as: A11, A12, A13, A14 indicators in the A1 indicator set), and compare the relative importance of the evaluation indicators in pairs to form a comparison matrix of evaluation indicators.

[0037] A = ( a ij ) = c 1 c 2 . . . c n c 1 a 11 a 12 . . . a 1 n c 2 a 21 a 22 . . . a 2 n . . . . . . . . . . . . . c n a n 1 a n 2 . . . a nn

[0038] a ij The value is selected using the three-scale method:

[0039] Table 1 Comparison matrix element determination

[0040]

[0041] Use the range method to construct the judgment matrix:

[0042] c ij = c b ( r i - r j ) / R

[0043] where: r i Is the sum of elements in each row of matrix A; C b is a constant, which is the relative importance of the extreme difference element pair given in advance according to a certain standard (generally, C is often used in practical applications) b =9). R=r max -r min , becomes extremely poor, r max =max{r 1 , r 2 ,...r n},r min =min{r 1 , r 2 ,...r n}. Matrix C = {c ij} n×n is the consistency judgment matrix.

[0044] Determine the weight of indicators at each level:

[0045] Find the largest eigenvalue λ of the judgment matrix C max And its corresponding feature vector W, that is, the weight vector W=(w 1 ,w 2 ,...w n ), by normalizing W, the relative weight of each evaluation index at a certain level with respect to its upper level index can be obtained.

[0046] w i ′ = w i / Σ i = 1 n w i , ( i = 1,2 , · · · , n )

[0047] Consistency check:

[0048] The weight value calculated by the AHP method needs to be checked for consistency. Calculate the largest eigenvalue λ max , and introduce the compatibility index CI to test the consistency of the judgment matrix.

[0049] CI=(λ max -n)/n-1

[0050] Generally, when CI<0.1, the judgment matrix is ​​considered to have satisfactory consistency; when CI≥0.1, the judgment matrix should be re-modified appropriately, and the consistency test should be performed again after correction.

[0051] (4) Correct the basic weight of the evaluation index.

[0052] In order to reduce the influence of many uncertain factors faced by the comparison matrix formed by a single expert, the joint decision-making of multiple experts is considered. Take X experts to participate in decision-making, and the weight vector formed by the judgment matrix made by k experts is (k=1,2,...,X), if the weight vector w (i) with w (j) The included angle is θ ij ,

[0053] S ij = cos ( θ ij ) = w ( i ) · w ( j ) | w ( i ) | | w ( j ) |

[0054] Then S ij Represents the judgment matrix C of the two experts (i) and C (j) the consistency between them. It can be seen that S ij The larger the value, the higher the consistency between the two judgment matrices, according to S ij Find the average consistency S of the i-th expert judgment matrix i.

[0055] S i = 1 X - 1 Σ j = 1 , j ≠ i X S ij

[0056] The average consistency of all experts is normalized to obtain the relative consistency S′ of the expert judgment matrix i , using the normalized relative consistency S′ i Multiplying with the corresponding weight vector and summing, the corrected weight can be obtained as:

[0057] W ′ = Σ i = 1 X S i ′ W ( i )

[0058] It can be seen that after the improvement, the weight values ​​obtained by different experts are treated differently, and the influence of relevant factors when different experts form the judgment matrix is ​​considered. Using it for weight revision can achieve better results and reduce the number of iterations.

[0059] (5) Comprehensive expert opinion evaluation and scoring.

[0060] The evaluation factor set F is a set composed of elements that affect the evaluation object, F={f1, f2,...fn}, the evaluation factor set in this paper is the 88 evaluation indicators of the evaluation system.

[0061] Evaluation factor value set V F V F ={V 1 , V 2 ,...V L}, where L is the number of evaluation values. This article selects V F ={90, 70, 50, 30, 10}, corresponding to {excellent, good, medium, poor, very poor} of the evaluation results, where L=5.

[0062] Based on the concept of fuzzy comprehensive evaluation, the mathematical model of fuzzy evaluation is determined as follows:

[0063] B=W'·R

[0064] Among them: W' is the weight vector, and R is the evaluation matrix.

[0065] Calculation of the evaluation matrix R: The evaluation matrix R is obtained by converting the expert opinion set. There are M known evaluation experts, and each expert scores the index from V F Set, taking the f1 index as an example, the evaluation score of the f1 index among M experts is V 1 for M 1 Personally, evaluate the f1 indicator score as V L for M L individual, the judgment matrix R1={M 1 /M, M 2 /M..., M L /M}, where: M=M 1 +M 2 …+M L.

[0066] Then the expert judgment matrix can be obtained:

[0067] R = M 11 / M M 12 / M . . . M 1 L / M M 21 / M M 22 / M . . . M 2 L / M . . . . . . M n 1 / M M n 2 / M . . . M nL / M

[0068] Calculation of comprehensive scoring results: normalize B until B', and compare B' with the evaluation factor value set V F Multiplying the score of the comprehensive evaluation can be obtained as:

[0069] G=B'·V T

[0070] The results of G refer to V F Specific values ​​can give specific evaluation results.