Efficiency evaluation algorithm based on cloud model

Abstract

The efficiency evaluation algorithm of the cloud model comprises the following steps: determining an evaluation index weight; constructing a cloud model, using a cloud generator tool, including a forward cloud generator and a reverse cloud generator, to realize mutual conversion between a qualitative concept and a quantitative value; an evaluation standard cloud and an evaluation cloud are generated, m evaluation indexes are scored for the n evaluation indexes by adopting an expert scoring method, and digital characteristics Ex, En and He are solved for the n indexes through a reverse cloud generator; a comprehensive evaluation cloud is calculated, and the index weights and the corresponding cloud digital features are calculated to solve the comprehensive evaluation cloud; and calculating the similarity with the standard clouds, calculating the similarity between the comprehensive evaluation clouds and the standard evaluation clouds after the comprehensive evaluation clouds are obtained, performing sorting according to the similarity, and taking the evaluation level corresponding to the standard cloud with the highest similarity as a final evaluation result. The method can effectively solve the problem that the uncertainty indexes are difficult to quantitatively evaluate in the field of efficiency evaluation.

Description

Technical Field

[0001] The present invention belongs to the technical field of engineering performance evaluation, and in particular to a performance evaluation algorithm based on a cloud model. Background Art

[0002] With the development of information technology and other high-tech technologies, foreign countries have begun to apply the research methods of system engineering to the research of engineering protection to solve the problem that only relying on hard technical protection means such as structural resistance under the threat of new technologies cannot ensure the survival and safety of position engineering. The US military was the first to introduce systematic thinking into the field of protection engineering, proposed the concepts of integrated protection and comprehensive protection, and developed a variety of optimization evaluation systems. The Soviet Union has also done a lot of work in this regard. They have conducted nearly 15 years of research, accumulated a lot of experience, and obtained many valuable research results, which have very important guiding significance for the updating of their protection means, the design and use of systems.

[0003] my country has conducted focused research in the field of engineering performance evaluation, mainly on the improvement or innovation of evaluation methods, including continuous analysis, induction and summary of the development laws and characteristics of performance evaluation, the connotation and extension of related concepts, establishment of an evaluation index system, rational selection of evaluation methods, exploration of the development direction of performance evaluation research and ways to deepen research, etc., and gradually forming a systematic theory to guide new performance evaluation research work.

[0004] However, in the existing field of engineering performance evaluation, there is still the problem that uncertainty indicators are difficult to quantify and evaluate. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a cloud model-based performance evaluation algorithm that can effectively solve the problem that uncertainty indicators in the performance evaluation field are difficult to quantify by introducing cloud theory to address the deficiencies of the prior art.

[0006] The technical problem to be solved by the present invention is achieved by the following technical solution. The present invention is a cloud model-based performance evaluation algorithm, comprising the following steps:

[0007] (1) Determine the weights of evaluation indicators;

[0008] (2) Building a cloud model

[0009] Use cloud generator tools, including forward cloud generator and reverse cloud generator, to convert between qualitative concepts and quantitative values;

[0010] (3) Generate evaluation standard cloud and evaluation cloud

[0011] The expert scoring method is used to give m evaluation indicators scores for n evaluation indicators, and the digital features E are obtained for n indicators through the reverse cloud generator. x 、E n and H e ;

[0012] (4) Calculation of comprehensive evaluation cloud

[0013] The indicator weights and the corresponding cloud digital features are calculated to obtain the comprehensive evaluation cloud;

[0014] (5) Calculate similarity with the standard cloud

[0015] After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud, sort them according to the similarity, and the evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result.

[0016] The technical problem to be solved by the present invention can also be further achieved by the following technical solution. For the above-mentioned cloud model-based performance evaluation algorithm, the specific process of step (1) is as follows:

[0017] For example, the calculation of the combined weight of each element on the kth layer relative to the highest layer depends on its relative weight relative to the layer above it, that is, the k-1th layer, and the combined weight of each element on the k-1th layer relative to the highest layer;

[0018] Assume that there are n elements on the k-1th layer. k-1 elements, the weight vector for the target layer, that is, the highest layer, is:

[0019]

[0020] Then the n on the kth layer k The weight vector of the element to the jth element on the previous layer, i.e. the k-1th layer, is:

[0021]

[0022] Then the matrix

[0023]

[0024] Yes k ×n k-1 The matrix of order represents the weight vector of the elements on the kth layer to the elements on the k-1th layer. Then the combined weight vector w of the elements on the kth layer to the target layer, that is, the highest layer, is k It can be calculated using the following formula:

[0025]

[0026] in,

[0027]

[0028] The technical problem to be solved by the present invention can also be further achieved by the following technical solution. For the above-mentioned cloud model-based performance evaluation algorithm, the specific process of step (2) is as follows:

[0029] Let U be a quantitative domain represented by exact numerical values. For any random value u∈U of the qualitative concept A, if we can find a random number μ with stable distribution characteristics A (u)∈[0,1], then μ A (u) is the membership of element u to concept A. The set of all memberships is called membership cloud, or cloud for short, namely:

[0030] The membership cloud C contains three digital features, namely the expected value E of the data represented by the membership cloud x 、Entropy E n and super entropy H e , which can be written as: C(E x , E n , H e );

[0031] Expected value E x : The expected value is the most representative numerical feature, which is the average value of all quantitative values ​​after the qualitative concept is converted. x The calculation formula is as follows:

[0032]

[0033] Entropy n : Entropy is a quantity used to measure the granularity of qualitative concepts and to indicate the degree of fuzziness of qualitative concepts. The greater the entropy, the more fuzzy the qualitative concept is and the greater the discreteness of its cloud map. Entropy E n The calculation formula is as follows:

[0034]

[0035] Super Entropy H e :Super entropy is an uncertain measure of entropy, which can also be called the entropy of entropy. It can better reflect the stability of the index, that is, it can reflect the degree of condensation of cloud droplets in the cloud map. Super entropy H e The calculation formula is as follows:

[0036]

[0037] Among them, S 2 is the sample variance, which is calculated as

[0038] The technical problem to be solved by the present invention can be further achieved by the following technical solutions. For the above-mentioned cloud model-based performance evaluation algorithm, the cloud generator is a tool for mutual conversion between qualitative and quantitative, including a forward cloud generator and a reverse cloud generator. Specifically:

[0039] The forward cloud generator is used to convert qualitative data into quantitative data. The input parameter feature E x , E n and H e And the number of cloud droplets N, output the value and membership of N cloud droplets, the specific algorithm flow is as follows:

[0040] (2.1.1) The expected value of the calculated input is E n , the variance is The normal random value E′ n ;

[0041] (2.1.2) The expected value of the calculated input is E x , the variance is E′ n 2 A normal random value x i ;

[0042] (2.1.3) According to the generated x i , E′ n Calculate the membership degree y i , the calculation formula is as follows:

[0043]

[0044] (2.1.4) According to x i and i Generate the corresponding cloud drop (x i ,y i );

[0045] (2.1.5) Repeat steps (2.1.1) to (2.1.4) until a sufficient number of cloud droplets are generated to terminate the algorithm;

[0046] The reverse cloud generator is used to complete the conversion from quantitative to qualitative. It inputs a number of cloud droplets and outputs three digital features of the cloud: x 、E n and H e , the specific algorithm flow is as follows:

[0047] (2.2.1) Calculate the sample mean of the quantitative data based on the algorithm input data

[0048] (2.2.2) Calculate the first-order center distance of the sample, that is, the expected value

[0049] (2.2.3) Calculating entropy

[0050] (2.2.4) Calculate the variance

[0051] (2.2.5) Calculate the super entropy based on the variance result. The calculation formula is as follows:

[0052]

[0053] (2.2.6) Get the digital feature E of the membership cloud corresponding to the qualitative concept x 、E n and H e .

[0054] The technical problem to be solved by the present invention can also be further achieved by the following technical solution. For the above-mentioned cloud model-based performance evaluation algorithm, the specific process of step (3) is as follows:

[0055] First, the performance is evaluated using four evaluation criteria: excellent, good, medium, and poor. The evaluation results are limited to the interval [0, 1]. Then, a cloud map is generated for each qualitative language value to establish a standard cloud evaluation scale.

[0056] Assume that the numerical range corresponding to the qualitative language is [B min , B max ], solve the cloud digital features of qualitative language according to the numerical interval of bilateral constraints, and generate cloud graph through forward cloud generator. The calculation formula is as follows:

[0057]

[0058] In the formula, k reflects the randomness of the evaluation value. The larger the value, the greater the randomness. Usually k=0.01. In this way, the evaluation standard cloud corresponding to excellent, good, medium and poor can be obtained.

[0059] The expert scoring method is used to give m evaluation indicators scores for n evaluation indicators, and the digital features E are obtained for n indicators through the reverse cloud generator. x 、E n and H e , the calculation formula is as follows:

[0060] Mean:

[0061] Expected value:

[0062] entropy:

[0063] variance:

[0064] Super Entropy:

[0065] Get the numerical features of n indicators: and

[0066] The technical problem to be solved by the present invention can also be further achieved by the following technical solution. For the above-mentioned cloud model-based performance evaluation algorithm, the specific process of step (4) is as follows:

[0067] The indicator weights and the corresponding cloud digital features are calculated to obtain the comprehensive evaluation cloud. The calculation formula is as follows:

[0068]

[0069]

[0070]

[0071] In the formula, n is the number of indicators, so we get the comprehensive evaluation cloud, which is recorded as: C(E x , E n , H e ).

[0072] The technical problem to be solved by the present invention can also be further achieved by the following technical solution. For the above-mentioned cloud model-based performance evaluation algorithm, the specific process of step (5) is as follows:

[0073] After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud, sort them according to the similarity, and the evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result;

[0074] Assume two cloud maps The cloud image similarity calculation method using N-point cloud drops is as follows:

[0075] (5.1) In cloud map C 1 China-Israel Generate N normal random numbers for the mean and variance

[0076] (5.2) In cloud map C 1 China-Israel Generate N normal random numbers x for the mean and variance i ;

[0077] (5.3) x i Substitute into cloud map C 2 The membership degree in the equation is:

[0078] (5.4) Calculate the cloud graph similarity:

[0079] The similarity of the cloud maps corresponding to excellent, good, medium and poor comprehensive evaluation clouds is calculated respectively. The evaluation value of the standard evaluation cloud corresponding to the highest similarity is the final evaluation value.

[0080] Compared with the prior art, the present invention can effectively solve the problem that uncertainty indicators in the field of performance evaluation are difficult to quantify by introducing cloud theory. Specifically, the present invention introduces a cloud model to conduct engineering performance evaluation. The cloud model is the basis of cloud-based computing, reasoning, and control. It is an uncertain conversion model for processing qualitative concepts and quantitative descriptions. It has been successfully applied to natural language processing, data mining, decision analysis, intelligent control, performance evaluation and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a flowchart of the algorithm steps of the present invention;

[0082] Figure 2 A digital characteristic graph of a cloud model of the present invention;

[0083] Figure 3 It is a schematic diagram of the evaluation criteria cloud of the present invention;

[0084] Figure 4 It is an image diagram of the comprehensive evaluation cloud model of the present invention;

[0085] Figure 5 It is a comparison diagram between the comprehensive evaluation cloud map of the present invention and the standard cloud map. DETAILED DESCRIPTION

[0086] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] Reference Figure 1-5 The present invention is a cloud model performance evaluation algorithm, which includes determining the evaluation index weight, building a cloud model, generating an evaluation standard cloud and an evaluation cloud, calculating a comprehensive evaluation cloud, and calculating the similarity with the standard cloud. Specifically:

[0088] (1) Determine the weights of evaluation indicators;

[0089] The specific process is:

[0090] For example, the calculation of the combined weight of each element on the kth layer relative to the highest layer depends on its relative weight relative to the layer above it (the k-1th layer), and the combined weight of each element on the k-1th layer relative to the highest layer;

[0091] Assume that there are n elements on the k-1th layer. k-1 elements, the weight vector for the target layer (the highest layer) is:

[0092]

[0093] Then the n on the kth layer k The weight vector of an element to the jth element on the previous layer (i.e., the k-1th layer) is:

[0094]

[0095] Then the matrix

[0096]

[0097] Yes k ×n k-1 The matrix of order represents the weight vector of the elements on the kth layer to the elements on the k-1th layer. Then the combined weight vector ω of the elements on the kth layer to the target layer (the highest layer) is k It can be calculated using the following formula:

[0098]

[0099] in,

[0100]

[0101] (2) Construct cloud models and use cloud generator tools, including forward cloud generator and reverse cloud generator, to achieve mutual conversion between qualitative concepts and quantitative values;

[0102] The specific process is:

[0103] Let U be a quantitative domain represented by exact numerical values. For any random value u∈U of the qualitative concept A, if we can find a random number μ with stable distribution characteristics A (u)∈[0,1], then μ A (u) is the membership of element u to concept A. The set of all memberships is called membership cloud, or cloud for short, namely:

[0104] The membership cloud C contains three digital features, namely the expected value E of the data represented by the membership cloud x 、Entropy E n and super entropy H e , which can be written as: C(Ex , E n , H e );

[0105] Expected value E x : The expected value is the most representative numerical feature and is the average of all quantitative values ​​after the qualitative concept is converted; the expected value E x The calculation formula is as follows:

[0106]

[0107] Entropy n : Entropy is a quantity used to measure the granularity of qualitative concepts. It is usually used to indicate the degree of fuzziness of qualitative concepts. The greater the entropy, the more fuzzy the qualitative concept is, and the greater the discreteness of its cloud map. Entropy E n The calculation formula is as follows:

[0108]

[0109] Super Entropy H e :Super entropy is an uncertain measure of entropy, which can also be called the entropy of entropy. It can better reflect the stability of the index, that is, it can reflect the degree of condensation of cloud droplets in the cloud map; super entropy H e The calculation formula is as follows:

[0110]

[0111] Among them, S 2 is the sample variance, which is calculated as

[0112] Cloud generator is a tool to achieve mutual conversion between qualitative and quantitative, including forward cloud generator and reverse cloud generator.

[0113] The forward cloud generator mainly realizes the conversion from qualitative to quantitative. The input parameter feature E x , E n and H e And the number of cloud droplets N, output the value and membership of N cloud droplets, the specific algorithm flow is as follows:

[0114] (2.1.1) The expected value of the calculated input is E n , the variance is The normal random value E′ n ;

[0115] (2.1.2) The expected value of the calculated input is E x , the variance is E′ n 2 A normal random value x i ;

[0116] (2.1.3) According to the generated xi , E′ n Calculate the membership degree y i , the calculation formula is as follows:

[0117]

[0118] (2.1.4) According to x i and i Generate the corresponding cloud drop (x i ,y i );

[0119] (2.1.5) Repeat steps (2.1.1) to (2.1.4) until a sufficient number of cloud droplets are generated to terminate the algorithm.

[0120] The reverse cloud generator mainly completes the conversion from quantitative to qualitative. It inputs a number of cloud droplets and outputs three digital features of the cloud: x 、E n and H e , the specific algorithm flow is as follows:

[0121] (2.2.1) Calculate the sample mean of the quantitative data based on the algorithm input data

[0122] (2.2.2) Calculate the first-order center distance of the sample, that is, the expected value

[0123] (2.2.3) Calculating entropy

[0124] (2.2.4) Calculate the variance

[0125] (2.2.5) Calculate the super entropy based on the variance result. The calculation formula is as follows:

[0126]

[0127] (2.2.6) Get the digital feature E of the membership cloud corresponding to the qualitative concept x 、E n and H e .

[0128] (3) Generate the evaluation standard cloud and the evaluation cloud. Use the expert scoring method to give m evaluation index scores for n evaluation indicators. Use the reverse cloud generator to obtain the digital features E for each of the n indicators. x E n and H e ;

[0129] The specific process is:

[0130] In the process of performance evaluation, there are inevitably many uncertain factors, which lead to the ambiguity and randomness of the evaluation indicators. Therefore, the performance can be evaluated by using four evaluation standards: excellent, good, medium, and poor. The evaluation results are limited to the interval [0, 1].

[0131] Qualitative language and intervals are not divided equally. The data intervals should be reasonably divided according to actual conditions and engineering experience so that they correspond one-to-one with the qualitative language. Expert consultation method can usually be used.

[0132] The correspondence between the assumed qualitative language and the distribution interval is shown in the table below.

[0133] Table 1 Correspondence between qualitative language and distribution interval

[0134]

[0135]

[0136] In this way, a cloud map can be generated for each qualitative language value, and a standard cloud evaluation scale can be established. Assume that the numerical range corresponding to the qualitative language is [B min , B max ], solve the cloud digital features of qualitative language according to the numerical interval of bilateral constraints, and generate cloud graph through forward cloud generator. The calculation formula is as follows:

[0137]

[0138] In the formula, k reflects the randomness of the evaluation value. The larger the value, the greater the randomness. Usually k = 0.01. In this way, the evaluation standard cloud corresponding to excellent, good, medium and poor can be obtained.

[0139] The expert scoring method is used to give m evaluation indicators scores for n evaluation indicators, and the digital features E are obtained for n indicators through the reverse cloud generator. x 、E n and H e , the calculation formula is as follows:

[0140] Mean:

[0141] Expected value:

[0142] entropy:

[0143] variance:

[0144] Super Entropy:

[0145] Get the numerical features of n indicators: and

[0146] (4) Calculate the comprehensive evaluation cloud, and solve the comprehensive evaluation cloud by calculating the indicator weights and the corresponding cloud digital features;

[0147] The specific process is:

[0148] The indicator weights and the corresponding cloud digital features are calculated to obtain the comprehensive evaluation cloud. The calculation formula is as follows:

[0149]

[0150]

[0151]

[0152] In the formula, n is the number of indicators; thus, the comprehensive evaluation cloud is obtained, which is recorded as: C(E x , E n , H e ).

[0153] (5) Calculate the similarity with the standard cloud. After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud and sort them according to the similarity. The evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result.

[0154] The specific process is:

[0155] After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud, sort them according to the similarity, and the evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result;

[0156] Assume two cloud maps The cloud image similarity calculation method using N-point cloud drops is as follows:

[0157] (5.1) In cloud map C 1 China-Israel Generate N normal random numbers for the mean and variance

[0158] (5.2) In cloud map C 1 China-Israel Generate N normal random numbers x for the mean and variance i ;

[0159] (5.3) x i Substitute into cloud map C 2 The membership degree in the equation is:

[0160] (5.4) Calculate the cloud graph similarity:

[0161] The similarity of the cloud maps corresponding to excellent, good, medium and poor comprehensive evaluation clouds is calculated respectively. The evaluation value of the standard evaluation cloud corresponding to the highest similarity is the final evaluation value.

Claims

1. A cloud model-based performance evaluation algorithm, characterized by: The following steps are involved: (1) Determine the weights of evaluation indicators; (2) Building a cloud model Use cloud generator tools, including forward cloud generator and reverse cloud generator, to convert between qualitative concepts and quantitative values; (3) Generate evaluation standard cloud and evaluation cloud The expert scoring method is used to give m evaluation indicators scores for n evaluation indicators, and the digital features E are obtained for n indicators through the reverse cloud generator. x 、E n and H e ; (4) Calculation of comprehensive evaluation cloud The indicator weights and the corresponding cloud digital features are calculated to obtain the comprehensive evaluation cloud; (5) Calculate similarity with the standard cloud After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud, sort them according to the similarity, and the evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result.

2. The cloud model-based performance evaluation algorithm according to claim 1, characterized in that: The specific process of step (1) is as follows: For example, the calculation of the combined weight of each element on the kth layer relative to the highest layer depends on its relative weight relative to the layer above it, that is, the k-1th layer, and the combined weight of each element on the k-1th layer relative to the highest layer; Assume that there are n elements on the k-1th layer. k-1 elements, the weight vector for the target layer, that is, the highest layer, is: Then the n on the kth layer k The weight vector of the element to the jth element on the previous layer, i.e. the k-1th layer, is: Then the matrix Yes k ×n k-1 The matrix of order represents the weight vector of the elements on the kth layer to the elements on the k-1th layer. Then the combined weight vector ω of the elements on the kth layer to the target layer, that is, the highest layer, is k It can be calculated using the following formula: in, 3. The cloud model-based performance evaluation algorithm according to claim 1, characterized in that: The specific process of step (2) is as follows: Let U be a quantitative domain U = {u} represented by exact numerical values. For any random value u∈U of the qualitative concept A, if we can find a random number μ with stable distribution characteristics A (u)∈[0,1], then μ A (u) is the membership of element u to concept A. The set of all memberships is called membership cloud, or cloud for short, namely: The membership cloud C contains three digital features, namely the expected value E of the data represented by the membership cloud x 、Entropy E n and super entropy H e , which can be written as: C(E x , E n , H e ); Expected value E x : The expected value is the most representative numerical feature, which is the average value of all quantitative values ​​after the qualitative concept is converted. x The calculation formula is as follows: Entropy n : Entropy is a quantity used to measure the granularity of qualitative concepts and to indicate the degree of fuzziness of qualitative concepts. The greater the entropy, the more fuzzy the qualitative concept is and the greater the discreteness of its cloud map. Entropy E n The calculation formula is as follows: Super Entropy H e :Super entropy is an uncertain measure of entropy, which can also be called the entropy of entropy. It can better reflect the stability of the index, that is, it can reflect the degree of condensation of cloud droplets in the cloud map. Super entropy H e The calculation formula is as follows: Among them, S 2 is the sample variance, which is calculated as 4. The efficiency evaluation algorithm based on the analytic hierarchy process according to claim 1 or 3 is characterized in that: The cloud generator is a tool for mutual conversion between qualitative and quantitative, including a forward cloud generator and a reverse cloud generator. Specifically: The forward cloud generator is used to convert qualitative data into quantitative data. The input parameter feature E x , E n and H e And the number of cloud droplets N, output the value and membership of N cloud droplets, the specific algorithm flow is as follows: (2.1.1) The expected value of the calculated input is E n , the variance is The normal random value E′ n ; (2.1.2) The expected value of the calculated input is E x , the variance is A normal random value x i ; (2.1.3) According to the generated x i , E′ n Calculate the membership degree y i , the calculation formula is as follows: (2.1.4) According to x i and i Generate the corresponding cloud drop (x i ,y i ); (2.1.5) Repeat steps (2.1.1) to (2.1.4) until a sufficient number of cloud droplets are generated to terminate the algorithm; The reverse cloud generator is used to complete the conversion from quantitative to qualitative. It inputs a number of cloud droplets and outputs three digital features of the cloud: x 、E n and H e , the specific algorithm flow is as follows: (2.2.1) Calculate the sample mean of the quantitative data based on the algorithm input data (2.2.2) Calculate the first-order center distance of the sample, that is, the expected value (2.2.3) Calculating entropy (2.2.4) Calculate the variance (2.2.5) Calculate the super entropy based on the variance result. The calculation formula is as follows: (2.2.6) Get the digital feature E of the membership cloud corresponding to the qualitative concept x 、E n and H e .

5. The performance evaluation algorithm based on the analytic hierarchy process according to claim 1 is characterized in that: The specific process of step (3) is as follows: First, the performance is evaluated using four evaluation criteria: excellent, good, medium, and poor. The evaluation results are limited to the interval [0,1]. Then, a cloud map is generated for each qualitative language value to establish a standard cloud evaluation scale. Assume that the numerical range corresponding to the qualitative language is [B min ,B max ], solve the cloud digital features of qualitative language according to the numerical interval of bilateral constraints, and generate cloud graph through forward cloud generator. The calculation formula is as follows: In the formula, k reflects the randomness of the evaluation value. The larger the value, the greater the randomness. Usually k=0.

01. In this way, the evaluation standard cloud corresponding to excellent, good, medium and poor can be obtained. The expert scoring method is used to give m evaluation indicators scores for n evaluation indicators, and the digital features E are obtained for n indicators through the reverse cloud generator. x 、E n and H e , the calculation formula is as follows: Mean: Expected value: entropy: Super Entropy: Get the numerical features of n indicators: and 6. The performance evaluation algorithm based on the analytic hierarchy process according to claim 1 is characterized in that: The specific process of step (4) is as follows: The indicator weights and the corresponding cloud digital features are calculated to obtain the comprehensive evaluation cloud. The calculation formula is as follows: In the formula, n is the number of indicators, so we get the comprehensive evaluation cloud, which is recorded as: C(E x ,E n ,H e ).

7. The performance evaluation algorithm based on the analytic hierarchy process according to claim 1 is characterized in that: The specific process of step (5) is as follows: After obtaining the comprehensive evaluation cloud, calculate its similarity with each standard evaluation cloud, sort them according to the similarity, and the evaluation level corresponding to the standard cloud with the highest similarity is the final evaluation result; Assume two cloud maps The cloud image similarity calculation method using N-point cloud drops is as follows: (5.1) In cloud diagram C1, Generate N normal random numbers for the mean and variance (5.2) In cloud diagram C1, Generate N normal random numbers x for the expected value and variance i ; (5.3) x i Substituting into the equation of cloud graph C2, the membership degree is: (5.4) Calculate the cloud graph similarity: The similarity of the cloud maps corresponding to excellent, good, medium and poor comprehensive evaluation clouds is calculated respectively. The evaluation value of the standard evaluation cloud corresponding to the highest similarity is the final evaluation value.

Images