A reactor state evaluation method based on game theory combined weight and improved topsis method

Abstract

The application discloses a kind of reactance state evaluation methods based on game theory combination weight and improved TOPSIS method, it is related to electrical equipment and mechanical equipment technical field, the subjective and objective weights of each evaluation index are calculated by analytic hierarchy process and entropy weight method respectively, the uncertainty and complexity in the operation data of the reactance of the evaluation method of the application are effectively handled, the cloud model theory is introduced, the correlation between each state index is quantitatively analyzed by expectation, entropy and hyper-entropy parameter, multi-dimensional operation data is converted into the comprehensive description of fuzzy and probability, this processing mode can more comprehensively capture the various characteristics of the operation state of the reactance, even in the case of uncertain data, a reasonable decision matrix can also be constructed, at the same time, the improved TOPSIS method introduces grey correlation degree to redefine relative closeness, further improve the accuracy and reliability of the evaluation result.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of electrical equipment and mechanical equipment, and particularly relates to an electric reactor state evaluation method based on a game theory combined weight and an improved TOPSIS method. BACKGROUND

[0002] The safe and stable operation of a power system is crucial to energy supply and economic development of a modern society, and an electric reactor, as an indispensable key device in the power system, directly determines the reliability and efficiency of the system. However, the state evaluation of the electric reactor is a complex multi-factor problem, and needs to comprehensively consider the influence of multiple indexes and multiple levels, and cope with inherent uncertainty and randomness in operation data, which makes the traditional evaluation method face multiple challenges in practical application, especially in the subjectivity of weight distribution, the deficiency in fuzzy processing ability and the limitation in multi-scheme optimization decision support.

[0003] The traditional electric reactor state evaluation method mainly relies on qualitative analysis and simple quantitative calculation, and the core idea is to evaluate the operation state of the electric reactor through single-threshold judgment of key indexes or empirical formula. These methods have the following limitations, including subjectivity of weight distribution: usually, expert scoring method or experience weighting method is adopted, which is too dependent on personal experience, so that the weight distribution result may deviate greatly from the actual situation; deficiency in multi-index comprehensive ability: the state of the electric reactor is affected by multiple indexes, including noise amplitude, oil test, oil chromatographic data and the like, and the traditional method is difficult to comprehensively consider the complex correlation between indexes, and is likely to ignore the potential coupling relationship; deficiency in fuzzy and random processing: the operation state indexes of the electric reactor usually have uncertainty, and the traditional method cannot effectively process the fuzziness and randomness in data, which may cause deviation of the evaluation result; single decision method: the traditional comprehensive evaluation methods such as weighted average method or simple TOPSIS method may have deficiencies in ideal solution distance processing, and are likely to cause unreasonable result sorting. The traditional method often cannot fully excavate the potential information in data, and is difficult to adapt to the high-precision requirement of modern complex power grid on equipment evaluation.

[0004] With the development of advanced theories and methods, the game theory has gradually become an important tool in the field of state evaluation due to its outstanding ability in multi-party decision and interest balance optimization, and the cloud model solves the problem of insufficient compatibility of traditional evaluation methods to fuzzy data and random data due to its significant advantages in fuzzy and random processing. SUMMARY

[0005] The purpose of the present application is to make up for the deficiencies of the prior art, and provide a reactor state evaluation method based on game theory combined weight and improved TOPSIS method, which can calculate the subjective and objective weights by using the analytic hierarchy process and entropy weight method respectively, and based on the game theory combination, the expert experience and the characteristics of the data itself are comprehensively considered, so that the weight distribution is more reasonable, the cloud model theory is introduced, the reactor operation data is converted into a comprehensive description of fuzzy and probability, and the uncertainty and complexity of the data are effectively handled, and the correlation between each state index is more accurately quantitatively analyzed.

[0006] The present application provides the following technical solutions to solve the above technical problems: a reactor state evaluation method based on game theory combined weight and improved TOPSIS method, the specific steps of the method are as follows:

[0007] S100, the subjective and objective weights of each evaluation index are calculated by using the analytic hierarchy process and entropy weight method respectively;

[0008] The analytic hierarchy process reflects the correlation degree of the reactor evaluation index from the qualitative and quantitative angles, and the importance of the index layer and the same layer elements is evaluated based on the historical data, and the evaluation matrix R is obtained by selecting the scale value A The evaluation matrix R is calculated A The characteristic vector alpha is obtained by solving, and the characteristic vector alpha is the weight coefficient, the consistency of the weight coefficient is checked, when the weight coefficient deviates from the actual operation state of the reactor, it is judged that the weight coefficient does not meet the requirements, and the evaluation matrix R is repeatedly solved A Until the characteristic vector meets the requirements, the weight of each index in the index layer is obtained;

[0009] The entropy weight method is used for multi-index evaluation and decision, the original data is processed by range standardization to construct the decision matrix X n , the entropy weight of each index is calculated to obtain the weight of each index;

[0010] The subjective and objective weights are optimized and combined based on the game theory method, and a complete weight distribution scheme is constructed;

[0011] S200, the qualitative concept theory of cloud model is introduced, the multi-dimensional operation state data of the reactor is converted into a comprehensive description of fuzzy and probability, the correlation between each state index is quantitatively analyzed by combining the operation data of the reactor, and then the evaluation decision matrix Z is constructed;

[0012] S300, the evaluation decision matrix Z established by the S200 is used to introduce the grey correlation degree in the TOPSIS method and redefine the relative closeness;

[0013] S400. Select the state level where the maximum proximity is located as the state of the reactor, and perform state assessment analysis using a reactor of model BKD-20000 / 330.

[0014] Furthermore, the evaluation matrix R constructed by S100 using the analytic hierarchy process (AHP) A for: Where r ij This represents the importance of the i-th element relative to the j-th element, where r ij and r ji They are reciprocals of each other.

[0015] Furthermore, S100 utilizes formula R A α=λ max α Solve for the evaluation matrix R A The eigenvector α in the λ, where α is the eigenvector and λ is the eigenvector. max The eigenvalue is the largest eigenvalue. The eigenvector is used as the weight coefficient. A consistency check is performed on the weight coefficient, and the consistency check uses the largest eigenvalue λ. max According to the consistency index n represents the number of indicators. The consistency ratio is calculated using the random consistency index RI. When CR ≤ 0.1, the consistency test is satisfied, and the subjective feature weight W1 = [w 11 ,w 12 ,…,w 1n ].

[0016] Furthermore, in multi-indicator evaluation and decision-making, S100 objectively calculates the weights of each indicator using the entropy weight method. The specific steps are as follows:

[0017] Construct an m×n decision matrix by setting up m decision options and n indicators.

[0018] The original data in the decision matrix is ​​subjected to range standardization, i.e., range standardization is applied. Where x minj and x maxj Let be the minimum and maximum values ​​of the j-th indicator;

[0019] Calculate the entropy value e of the j-th index. j for: in

[0020] Calculate the entropy weight w of the j-th index. j for: Calculate the entropy weights of all indicators to obtain the objective feature weights W2 = [w 21 ,w 22 ,…,w2n ].

[0021] Furthermore, the qualitative concept of the cloud model in S200 is derived from the expectation E. x Entropy E n and hyperentropy H e To represent it as a whole, that is, Y(E) x E n H s ), wherein the expected value E x Let E be the center point of the universe of discourse for the cloud droplets in the cloud model, representing the average value of the cloud droplets in the universe of discourse, to reflect the characteristic quantity of the insulation state of the equipment under this index. n The hyperentropy H represents the width of the expected cloud curve and measures the uncertainty of things in the universe of discourse. e It is a measure of E n The measure of uncertainty, namely the entropy of entropy, reflects both the dispersion of entropy and the degree of cloud droplet condensation. H e A larger value indicates a thicker cloud layer in the cloud model. The calculation process for the digital features of the cloud model is as follows:

[0022] Double-constrained space [c min c max ] represents the boundary of the insulation state index for each level, then the expected value E of the cloud model for each level is... x The average value is: c min and c max These represent the critical minimum and maximum values, respectively.

[0023] Adopting "3E" n The calculation rule for "will not belong to the interval [E]" x -3E n E x +3E n The cloud droplets are considered low-probability events, meaning they cannot occur and do not affect the overall characteristics of the cloud models at each level. Therefore, E... n The value can be:

[0024] via E x Given the index value x, calculate the degree of certainty that x belongs to the cloud model. And based on the values ​​of each indicator and the certainty k of the corresponding cloud model, the evaluation decision matrix Z is constructed as follows: Where m is the number of evaluation indicators, n is the number of status levels, and Z mn Let m be the correlation degree of the m-th evaluation index at the n-th state level. The rows of the matrix represent different evaluation indices, and the columns represent different state levels.

[0025] Further, the S300 introduces the grey correlation degree in the TOPSIS method to redefine the relative closeness, and determines the positive and negative ideal solutions as:

[0026] The distance between each alternative solution and the positive and negative ideal solutions is calculated as:

[0027] The grey correlation degree coefficient of each alternative solution to the positive and negative ideal solutions is calculated as: wherein, and is the grey correlation coefficient, and ρ∈(0, 1) is the resolution coefficient. When ρ≈1, the difference between the correlation coefficients is small, and the distinguishing ability is weak. When ρ≈0, the difference is large, and the distinguishing ability is strong. When ρ=0.5,

[0028] The grey correlation degrees of each alternative solution under different attributes are averaged to obtain the grey correlation degree of the alternative solution.

[0029] The is normalized respectively, and a new relative closeness is obtained by fusion, that is, wherein η represents the preference degree of the decision maker to the distance of the positive and negative ideal solutions, and κ represents the preference degree of the decision maker to the grey correlation degree, satisfying η+κ=1 and η=0.5, κ=0.5.

[0030] The relative closeness is: According to the ranking result of the relative closeness, the optimal selection of the alternative solution is performed, and the relative closeness S i ≈1, indicating that the evaluation result of the alternative solution is excellent, and vice versa.

[0031] Further, when the S400 performs state evaluation analysis through the electric reactor with the model of BKD-20000 / 330, the step of designing the evaluation analysis network is:

[0032] The subjective weight of each index is obtained by the S100, the objective weight of each index is obtained, and the combination weight is obtained by optimizing the subjective and objective weights;

[0033] The correlation degree of the evaluation index under each state level is calculated by the grade cloud model, and the decision evaluation matrix is constructed by the correlation degree;

[0034] The positive ideal solution the negative ideal solution and the grey correlation degree Calculate the correlation degree under each state level according to the relative closeness;

[0035] Select the state level where the maximum closeness is located as the state of the reactor, and formulate the corresponding operation and maintenance strategy.

[0036] Compared with the prior art, the reactor state evaluation method based on game theory combined weight and improved TOPSIS method has the following beneficial effects:

[0037] First, the evaluation method effectively handles the uncertainty and complexity of the reactor operation data, introduces the cloud model theory, quantitatively analyzes the correlation degree between each state index through the expectation, entropy and hyper entropy parameters, and converts the multi-dimensional operation data into a comprehensive description of fuzzy and probability. This processing method can more comprehensively capture various characteristics of the reactor operating state. Even in the case of uncertain data, a reasonable decision matrix can be constructed. At the same time, the improved TOPSIS method introduces the grey correlation degree to redefine the relative closeness, further improving the accuracy and reliability of the evaluation results.

[0038] Second, the reactor state evaluation method can more accurately determine the weight of the evaluation index. The analytic hierarchy process reflects the index correlation degree from qualitative and quantitative angles, and the consistency test ensures that the weight is reasonable. The entropy weight method objectively calculates the weight to avoid subjective bias. Combined with game theory, the weights of the two are combined, which comprehensively considers subjective and objective factors, making the weight distribution more scientific. This reasonable weight determination method can better reflect the influence degree of each index on the state of the reactor, and avoid evaluation deviation caused by unreasonable weight.

[0039] Other advantages, objects and features of the present application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, without creative labor, other drawings can also be obtained from these drawings.

[0041] Figure 1 It is a flow chart of the reactor state evaluation method based on game theory combined weight and improved TOPSIS method;

[0042] Figure 2 It is a flow chart of S300 in the TOPSIS method to redefine the relative closeness. DETAILED DESCRIPTION

[0043] To further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific embodiments, structures, features and effects thereof according to the present application are described in detail below in conjunction with the drawings and preferred embodiments.

[0044] Example One

[0045] This embodiment details the application of a reactor state evaluation method based on game theory combined weights and improved TOPSIS method in the daily operation and maintenance of power systems. The subjective and objective weights are calculated by the analytic hierarchy process and entropy weight method respectively and optimized combination, the cloud model theory is introduced to construct the decision matrix and the grey correlation degree to redefine the relative closeness, and finally the reactor state is determined and the corresponding operation and maintenance strategy is developed. This method effectively overcomes the limitations of traditional evaluation methods and improves the accuracy and reliability of reactor state evaluation.

[0046] In the specific implementation, first, an evaluation matrix is constructed, the weight is calculated by the analytic hierarchy process, and the correlation degree of the reactor evaluation index is reflected from the qualitative and quantitative angles. Based on the historical data, the importance of the elements in the same layer of the index layer is evaluated, the scale value is selected to construct the evaluation matrix R A Because there is a complex correlation between the various indexes of the reactor, the relationship is quantified by the scale value, which can provide a basis for subsequent weight calculation. The evaluation matrix R A is constructed where r ij represents the importance of the i-th element relative to the j-th element, and r ij and r ji are inverses of each other. To ensure the symmetry of the evaluation, the importance evaluation of the index is more reasonable. The eigenvector is solved and consistency check is performed. The characteristic vector a in the evaluation matrix R A is solved by the formula R max α=λ A , where a is the eigenvector, λ max is the maximum eigenvalue, and the eigenvector is used as the weight coefficient. The eigenvector can reflect the relative importance of each index in the overall evaluation. The consistency check of the weight coefficient is to ensure that the importance relationship of the index reflected by the evaluation matrix is reasonable, avoiding logical contradictions. According to the consistency index n is the number of indexes, and the consistency ratio is calculated in combination with the random consistency index RI When CR≤0.1, the consistency check is satisfied, and the subjective characteristic weight W1=[w 11 ,w 12 ,…,w 1nIf the consistency check fails, it means that the evaluation matrix is not reasonable and the importance evaluation between indexes needs to be adjusted until the consistency check requirement is met.

[0047] Then, for multi-index evaluation and decision-making, m decision schemes and n indexes are set to construct an m x n decision matrix X n For comprehensive evaluation of multiple indexes, considering the different value ranges and properties of different indexes, the original data is standardized by range standardization, that is, range standardization is used Where x minj and x maxj are the minimum and maximum values of the jth index, respectively. Through this standardization, the values of each index are mapped to a relatively unified interval, facilitating subsequent calculation of entropy values and entropy weights. For calculation of the entropy value e j of the jth index, the formula is Where The entropy value reflects the dispersion degree and uncertainty of the index data. The calculated entropy value further determines the weight of each index in the comprehensive evaluation. For calculation of the entropy weight w j of the jth index, the formula is The entropy weights of all indexes are calculated to obtain the objective feature weight W2 = [w 21 , w 22 , …, w 2n ]. The greater the entropy weight, the higher the importance of the index in the comprehensive evaluation, reflecting the degree of uncertainty of the index data. The higher the uncertainty, the greater the weight of the index in the evaluation.

[0048] Subsequently, based on the game theory combination weight, the subjective weight vector obtained by the analytic hierarchy process is W1 = [w 11 , w 12 , …, w 1n ], and the objective weight vector obtained by the entropy weight method is W2 = [w 21 , w 22 , …, w 2n ]. According to the game theory method, the combination weight vector W can be calculated by the formula W = αW1 + (1-α)W2, where α is the allocation proportion of the weight vector. The subjective and objective factors are considered to affect the index weight. The subjective weight reflects the expert experience and qualitative judgment of the importance of the index, and the objective weight is based on the actual situation of the data. Through combination of weights, a more reasonable and accurate weight allocation scheme can be obtained, making the final evaluation result more consistent with the actual situation. Then, the decision evaluation matrix is constructed, and the qualitative concept of the cloud model is used to obtain the expectation E x , entropy E nand hyperentropy H e To represent it as a whole, that is, Y(E) x E n H s ), where the expected value is E x Here, represents the center point of the universe of discourse for cloud droplets in the cloud model, indicating the average value of cloud droplets within the universe of discourse, thus reflecting a characteristic quantity of the insulation state of the equipment under this index; Entropy E n In the universe of discourse, it represents the width of the expected cloud curve and measures the uncertainty of things; hyperentropy H e It is a measure of E n The measure of uncertainty, namely the entropy of entropy, reflects both entropy and the degree of cloud droplet condensation, H. e The larger the value, the thicker the cloud layer in the cloud model. The cloud model theory is introduced because the reactor operating data has uncertainty and complexity. The cloud model can effectively handle this fuzziness and randomness, transforming the multi-dimensional operating state data of the reactor into a comprehensive description of fuzziness and probability. Digital feature calculations are performed on the cloud model, utilizing the double-constraint space [c min ,c max ] represents the boundary of the insulation state index for each level, then the expected value E of the cloud model for each level is... x The average value is c min and c max These represent the minimum and maximum critical values, respectively. This selection is based on the characteristics of the double-constraint space, aiming to use the midpoint of the interval to better represent the characteristics of that level, allowing the cloud model to more accurately describe the state of the indicator. This is achieved using a "3E"-based approach. n The calculation rule for "will not belong to the interval [E]" x -3E n E x +3E n If cloud droplets are considered low-probability events, meaning events that cannot occur and do not affect the overall characteristics of cloud models at each level, then E n Values This value selection method is based on reasonable control of the uncertainty of the cloud model. By setting an entropy value, the cloud model can reflect the uncertainty of the indicators to a certain extent, while preventing the model from losing accuracy due to excessive uncertainty. An evaluation decision matrix is ​​constructed, and E... x Given the index value x, calculate the degree of certainty that x belongs to the cloud model. And based on the values ​​of each indicator and the certainty k of the corresponding cloud model, the evaluation decision matrix Z is constructed as follows: Where m is the number of evaluation indicators, n is the number of status levels, and Z mnThe correlation degree of the mth evaluation index in the nth state level is obtained by constructing the evaluation decision matrix, the rows of the matrix represent different evaluation indexes, and the columns represent different state levels. The evaluation decision matrix can quantitatively analyze the correlation between the state indexes, and provide a data basis for subsequent improvement of the TOPSIS algorithm. By calculating the degree of certainty k, the matching degree of the index value and the cloud model can be measured, so that the relationship between the index in different state levels can be better reflected.

[0049] Next, the improved TOPSIS algorithm is applied. The correlation degree of each evaluation index of the reactor in different state levels is obtained from the constructed evaluation decision matrix Z. The grey correlation degree is introduced to redefine the relative closeness of the positive and negative ideal solutions as follows: The positive and negative ideal solutions are determined to measure the closeness of the alternative scheme to the ideal state. The positive ideal solution represents the optimal value of each index in the corresponding level, and the negative ideal solution represents the worst value of each index in the corresponding level. By comparing the distance and correlation degree between the alternative scheme and the positive and negative ideal solutions, the advantages and disadvantages of the alternative scheme are evaluated. The distance to the positive and negative ideal solutions is calculated, and the distance between each alternative scheme and the positive and negative ideal solutions is calculated as follows: The positive and negative ideal distance can quantitatively measure the spatial distance between the alternative scheme and the ideal solution. The smaller the distance, the closer the alternative scheme to the ideal solution. Then the grey correlation degree coefficient and the grey correlation degree of each alternative scheme to the positive and negative ideal solutions are calculated as follows: wherein p ∈ (0, 1) is a resolution coefficient. When p ≈ 1, the difference between the correlation coefficients is small, and the distinguishing ability is weak. When p ≈ 0, the difference is large, and the distinguishing ability is strong. When p = 0.5, The grey correlation degree coefficient and the grey correlation degree are calculated to measure the relationship between the alternative scheme and the ideal solution from the perspective of correlation degree. The grey correlation degree coefficient reflects the correlation degree of the alternative scheme and the ideal solution in each index. The grey correlation degree is obtained by calculating the average value, which can comprehensively evaluate the correlation between the alternative scheme and the ideal solution. The are normalized to obtain a new relative closeness, that is: wherein η represents the preference degree of the decision maker to the distance of the positive and negative ideal solutions, and κ represents the preference degree of the decision maker to the grey correlation degree. η + κ = 1 and η = 0.5, κ = 0.5 are satisfied. The new relative closeness considers the influence of distance and correlation degree on the evaluation of the alternative scheme, so that the evaluation result is more comprehensive and accurate. According to the ranking result of the relative closeness, the optimal selection of the alternative scheme is made, and the relative closeness The relative closeness is compared to determine the advantages and disadvantages of the alternative scheme, and the relative closeness S i≈1, it indicates that the evaluation result of the alternative is excellent, otherwise the evaluation result is poor.

[0050] Finally, the state grade where the maximum closeness degree is located is selected as the state of the reactor. By comparing the relative closeness degrees under each state grade, the state grade corresponding to the maximum closeness degree is determined, which is the actual state of the reactor. According to the state of the reactor, a corresponding operation and maintenance strategy is formulated, that is, if the reactor is in good condition, the maintenance period is appropriately extended, and the normal operation of the power system is not affected; if it is in poor condition, maintenance is arranged in time to avoid failure and ensure the safe and stable operation of the power system.

[0051] To sum up, the embodiment introduces in detail the application of the reactor state evaluation method based on game theory combined weight and improved TOPSIS method in the daily operation and maintenance of the power system. By accurately calculating the weight of each evaluation index, reasonably constructing the decision matrix, and accurately applying the improved TOPSIS algorithm, the state of the reactor can be accurately determined and the corresponding operation and maintenance strategy can be formulated.

[0052] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, without departing from the scope of the technical solution of the present application. Any simplification, modification, equivalent change and modification of the above embodiments made according to the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

1. A reactor state evaluation method based on game theory combined weight and improved TOPSIS method, characterized in that, Specific steps of the method are as follows: S100, calculating subjective and objective weights of each evaluation index by analytic hierarchy process and entropy weight method respectively; The analytic hierarchy process reflects the correlation degree of the reactor evaluation indexes from the qualitative and quantitative aspects, evaluates the importance of the elements in the index layer based on historical data, and obtains an evaluation matrix R by selecting a scale value A , the evaluation matrix R A is solved to obtain an eigenvector α, the eigenvector α is a weight coefficient, the weight coefficient is subjected to a consistency test, when the weight coefficient deviates from the actual operation state of the reactor, it is determined that the weight coefficient does not meet the requirements, and the evaluation matrix R is repeatedly solved A until the eigenvector meets the requirements, and the weight of each index in the index layer is obtained. The entropy weight method is used for multi-index evaluation and decision, and a decision matrix X is constructed by performing range standardization on original data n , and entropy weight of each index is calculated to obtain index weight; Combining the subjective and objective weights based on game theory to construct a complete weight distribution scheme; S200, introducing the qualitative concept theory of cloud model, converting the multi-dimensional running state data of the reactor into a comprehensive description of fuzzy and probability, combining the running data of the reactor, quantitatively analyzing the correlation degree between each state index, and then constructing an evaluation decision matrix Z; S300, introducing the grey correlation degree into the TOPSIS method and redefining the relative closeness degree by using the evaluation decision matrix Z established in S200; S400, selecting the state grade where the maximum closeness degree is located as the state of the reactor, and performing state evaluation analysis on the reactor with model BKD-20000 / 330; Steps of the state evaluation analysis network are as follows: S100, obtaining the subjective weight of each index, obtaining the objective weight of each index, and optimizing the subjective and objective weights to obtain the combined weight; S200, calculating the correlation degree of the evaluation index under each state grade by the grade cloud model, and constructing a decision evaluation matrix by the correlation degree; The positive ideal solution under each state level is obtained from the decision matrix The negative ideal solution And the grey correlation degree The correlation degree under each state level is calculated according to the relative closeness degree; S300, selecting the state grade where the maximum closeness degree is located as the state of the reactor, and formulating the corresponding operation and maintenance strategy.

2. The method according to claim 1, wherein, The S100 is constructed by the analytic hierarchy process evaluation matrix R A For: Where r ij Represents the importance of the i-th element relative to the j-th element, where r ij And r ji Are reciprocal.

3. The method of claim 1, wherein the method is characterized by, The S100 utilizes the formula R A α = λ max Solving the evaluation matrix R A characteristic vector α in, wherein α is a characteristic vector, λ max is the maximum eigenvalue, taking the characteristic vector as a weight coefficient, and performing a consistency test on the weight coefficient, wherein the consistency test utilizes the maximum eigenvalue λ max , and the consistency index n is the number of indexes, and a consistency ratio CR is calculated in combination with a random consistency index RI When CR≤0.1, the consistency test is satisfied, and a subjective characteristic weight W1 = [w 11 ,w 12 ,…,w 1n ] is obtained.

4. The method of claim 1, wherein the method is characterized by, In the multi-index evaluation and decision, the entropy weight method is used to objectively calculate the weight of each index, and the specific steps are as follows: An m decision scheme and n indexes are set to build an m x n decision matrix The original data in the decision matrix is processed by range standardization, i.e. range standardization is adopted Wherein x minj and x maxj are the minimum value and maximum value of the jth index calculating an entropy value e of the jth indicator j is: wherein Calculate the entropy weight w of the jth index j is: Calculate the entropy weight of all indexes to get the objective feature weight W2 = [w 21 , w 22 , …, w 2n ].

5. The method of claim 1, wherein the method is characterized by, The qualitative concept of the cloud model in S200 is represented by expectation E x , entropy E n , and hyper-entropy H e as a whole, that is, Y(E x , E n , H s ), wherein the expectation E x is the center point of the domain of the cloud drop of the cloud model, which represents the average value of the cloud drop in the domain space, so as to embody the characteristic quantity of the insulation state of the equipment under this index, the entropy E n represents the width of the expected cloud curve in the domain space and measures the uncertainty of things, and the hyper-entropy H e is a measure of the uncertainty of E n , that is, entropy of entropy, which reflects both the dispersion degree of entropy and the condensation degree of the cloud drop, and the greater the value of H e , the thicker the cloud layer in the cloud model, and for the calculation of the digital features of the cloud model, the calculation process is: Double-constrained space [c min , c max ] represents the boundary of each level insulation state index, then the expectation E x of each level cloud model is valued as: c min and c max represent the minimum and maximum values of the critical point respectively; Using the calculation rule based on "3E n ", the cloud drops not belonging to the interval [E x -3E n , E x +3E n ] are considered as a small probability event, i.e. cannot occur and do not affect the overall characteristics of each level cloud model, then E n is valued as: By E x and the index value x, the degree of certainty that x belongs to the cloud model is calculated and according to the degree of certainty k of each index value and the corresponding cloud model, an evaluation decision matrix Z is constructed as: where m is the number of evaluation indexes, n is the number of state levels, Z mn is the correlation degree of the mth evaluation index under the nth state level, the rows of the matrix represent different evaluation indexes, and the columns represent different state levels.

6. The method of claim 1, wherein the method is characterized by, The S300 introduces the grey correlation degree in the TOPSIS method to redefine the relative closeness, and determines the positive and negative ideal solutions as follows: The distance between each alternative to the positive and negative ideal solutions is calculated as: The grey correlation degree coefficients of each alternative to the positive and negative ideal solutions are calculated respectively as follows: wherein, and is a grey correlation coefficient, and ρ ∈ (0, 1) is a resolution coefficient. When ρ ≈ 1, the difference between the correlation coefficients is small, and the distinguishing ability is weak. When ρ ≈ 0, the difference is large, and the distinguishing ability is strong. When ρ = 0.5, The grey correlation degrees of the alternatives are obtained by averaging the grey correlation coefficients of each alternative under different attributes. The normalized processing is respectively performed on to obtain a new relative closeness degree, that is wherein η represents the preference degree of the decision maker to the distance between the positive ideal solution and the negative ideal solution, κ represents the preference degree of the decision maker to the grey correlation degree, and satisfies η+κ=1 and η=0.5, κ=0.

5. The relative closeness is: The optimal selection of the alternative is made according to the ranking result of the relative closeness, and the relative closeness S i ≈1, which indicates that the evaluation result of the alternative is excellent, and otherwise, the evaluation result is poor.

Images