A modernization evaluation method for embankment projects based on a combination of cloud model and entropy weight

Abstract

This invention discloses a method for evaluating levee project modernization based on a combination of cloud models and entropy weights. The method comprises: constructing an evaluation index system for levee project management modernization; converting evaluation index data into a cloud model based on the constructed levee project management modernization evaluation index system; determining the combined weights of various indicators through combined weighting based on the cloud model, entropy weighting, and game theory; constructing a levee project management modernization evaluation model based on cloud theory; and conducting specific evaluations using the levee project management modernization evaluation model based on cloud theory. This invention clarifies the connotation and content of levee project management modernization, constructs a multi-level evaluation index system, and establishes an evaluation model based on a combination of cloud models and entropy weights, providing a more comprehensive and scientific evaluation model for levee management modernization.

Description

Technical Field

[0001] The present invention relates to a levee engineering modernization evaluation technology, and in particular to a levee engineering modernization evaluation method based on a combination of a cloud model and entropy weight. Background Art

[0002] The modernization of embankment engineering management is a dynamic process of establishing a comprehensive, modern, standardized, professional and refined embankment engineering management system against the background of sustained and rapid development of contemporary social economy. It is a systematic project that includes management systems, mechanisms, means, talents, etc.

[0003] my country has initially established a comprehensive levee management system that allows for real-time monitoring of water and rainfall conditions, monitoring of construction conditions, risk analysis and management, and information sharing. Numerous studies have examined the necessity, existing challenges, and countermeasures for levee management modernization. However, research on levee management modernization and its evaluation is still in its infancy. The hierarchy and comprehensiveness of the indicator system need to be expanded, and most evaluation attributes are difficult to quantify. It is necessary to refine the levee management modernization indicator system based on the current characteristics and actual conditions of levee management in my country, fully incorporating research findings from water conservancy modernization and water conservancy project management modernization, and select appropriate evaluation methods and models for comprehensive multi-indicator evaluation. Summary of the Invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a method for evaluating the modernization of embankment projects based on a combination of cloud model and entropy weight, to apply the combination of cloud model and entropy weight to the evaluation of embankment project modernization, and to establish a more comprehensive and scientific evaluation model for the field of embankment project modernization evaluation.

[0005] Technical solution: The present invention provides a method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight, comprising the following steps:

[0006] S1. Construct a modern evaluation index system for embankment project management;

[0007] S2. According to the constructed embankment engineering management modernization evaluation index system, the evaluation index data is converted into the cloud model; wherein, the cloud model uses three independent parameters: expectation Ex, entropy En, and super entropy H. e To express a qualitative concept; evaluation index data includes quantitative index data and qualitative index data. T evaluation values ​​of a quantitative index in different states are integrated into a cloud model to represent it; qualitative indicators are set with comment levels according to needs. According to the correspondence between the level comments and value ranges of qualitative indicators, the qualitative indicator data and the cloud model are converted;

[0008] S3. Determine the combined weights of various indicators based on the combined weighting of the cloud model, entropy weight method, and game theory. First, determine the weight distribution results of each indicator based on expert theory, calculate the weight cloud model parameters of each indicator, perform accuracy test and parameter correction on the weight cloud model, calculate the weight calculation value of each indicator in the corrected weight cloud model, and normalize the weight calculation values ​​of all indicators to obtain a weight set. Secondly, obtain the cloud model of each indicator according to the method in step S2, calculate the entropy weight of each indicator, form a weight vector, and obtain the final objective weight. Finally, perform any linear combination of multiple weight vectors in the weight vector set to obtain a feasible weight vector, and find the most suitable combination weight in the feasible vector set according to game theory.

[0009] S4. Construct a cloud theory-based evaluation model for the modernization of levee project management. First, collect relevant information and obtain cloud models for each indicator according to the method in step S2. Second, replace the qualitative comments with the scores of the qualitative indicator cloud model, form a quantitative evaluation matrix with the original quantitative indicator data, and perform weighted calculation with the indicator combination weights obtained in step S3. Repeat the operation nt times to obtain a secondary indicator evaluation cloud model. Finally, calculate the secondary indicator evaluation cloud model to obtain the overall goal realization cloud model and the primary indicator realization cloud model, and compare them with the scale cloud model of the evaluation grade standard to determine the evaluation grade of the overall goal modernization realization degree and the evaluation grade of the primary indicator modernization realization degree.

[0010] S5. Use the cloud theory-based modern evaluation model for embankment engineering management to conduct specific evaluation.

[0011] Furthermore, step S1 is specifically as follows:

[0012] S11. Use frequency analysis to cluster and summarize the indicators listed in existing relevant documents and literature research, and conduct forward-looking expansion;

[0013] S12. Use theoretical analysis to select representative indicators based on the components of the evaluation object and their internal connections;

[0014] S13. The Delphi method was used to consult experts and the principal component analysis method was used to optimize the indicators and construct an evaluation index system for embankment engineering management modernization, which includes five first-level indicators, 22 second-level indicators, and 99 third-level indicators.

[0015] S14. Based on the constructed embankment engineering management modernization evaluation index system, determine the corresponding relationship between the grade comments and value ranges of all qualitative indicators except the five quantitative indicators: the cleaning rate within the demarcation area reaches 50%, the cleaning rate within the title confirmation area reaches 90%, the greening coverage rate within the demarcation area reaches 50%, the greening coverage rate within the title confirmation area reaches 90%, and the soil and water loss control rate reaches 90%, as shown in Table 1:

[0016] Table 1 Index grade comments and their corresponding score ranges

[0017]

[0018] Furthermore, the method for establishing the quantitative indicator data and cloud model conversion model in step S2 is:

[0019] The t evaluation values ​​of a quantitative indicator in different states can be integrated into a cloud model to represent it, and the expected Ex, entropy En, and super entropy H are obtained by using the reverse cloud generator. e The calculation formula is:

[0020]

[0021]

[0022]

[0023]

[0024] Among them, Ex is the expectation of the cloud model; En is the entropy of the cloud model; H e is the super entropy of the cloud model; x i is the i-th evaluation value of the quantitative indicator; S 2 is the variance of the quantitative indicator; t is the number of evaluation values ​​of the quantitative indicator;

[0025] Or it can be obtained by simplifying the calculation using the following formula. In this case, the quantitative index corresponds to the super entropy H of the cloud model: e is a specified constant;

[0026]

[0027] Furthermore, the method for establishing the qualitative indicator data and cloud model conversion model in step S2 is:

[0028] Qualitative indicators are set with different numbers and actual meanings of evaluation levels according to needs. The domain corresponding to the evaluation level of qualitative indicators is Ω, and Ω′ is a division of Ω:

[0029]

[0030] Among them, Ω i is the domain corresponding to the i-th level of comment; k is the number of comment levels;

[0031] The normal evaluation cloud model of the i-th level qualitative comment grade on the partition domain is called the grade evaluation cloud model, and its digital feature is T i =(Ex i ,En i ,Hei ); T1 and T k When they are defined as half falling cloud and half rising cloud respectively, the evaluation value interval of level i is [a i ,b i ], where i = 1, 2, ..., k, based on the 3En rule of the normal cloud model, the cloud parameter of the i-th level (Ex i ,En i ) According to the following formula, the super entropy He i is a constant;

[0032]

[0033] According to the corresponding relationship between the grade comments and value ranges of the qualitative basic indicators, the qualitative indicator data and the cloud model are converted according to the above formula to obtain the cloud model parameters of the comment grade;

[0034] When an indicator of the evaluation object has multiple grade comment values, the following formula is used to calculate the expected Ex, entropy En, and super entropy H of the grade evaluation cloud corresponding to the indicator. e ;

[0035]

[0036] Where H is the number of grade evaluation values; Ex i′ is the expectation of the grade evaluation cloud corresponding to the i′th grade comment value; En i′ is the entropy of the grade evaluation cloud corresponding to the i′th grade comment value; He i′ is the super entropy of the grade evaluation cloud corresponding to the i′th grade comment value.

[0037] Furthermore, step S3 is specifically as follows:

[0038] S31. Invite n experts to give the weight distribution results of m indicators: Where i1 = 1, 2, ..., m, j = 1, 2, ..., n; using the reverse cloud generator that does not require the degree of certainty μ(x), the weighted cloud model parameters are obtained by calculating the n weighted evaluation values ​​of the indicator i1 The weighted cloud model of index i1 is tested for accuracy and its parameters are corrected. The forward cloud algorithm is used to randomly generate q cloud droplets for the corrected weighted cloud model of index i1, and the score corresponding to the maximum certainty of the cloud droplet point is used as the weight calculation value ω of index i1. i ; Normalize the weight calculation values ​​of m indicators to obtain the weight set:

[0039] S32. For each qualitative grade comment of the indicator, transform it according to step S2 to obtain cloud model parameters; use the forward cloud algorithm to generate a random cloud droplet for each cloud model and replace it with the qualitative comment in the evaluation result matrix; when there are w groups of original data for m evaluation indicators, the evaluation indicator eigenvalue matrix is: Where j1 = 1, 2, ..., w; the eigenvalue matrix is ​​normalized to obtain the relative superiority matrix R = (ri1j1) m×w ,in is the normalized value of the indicator; calculate the entropy value is the normalized index value; the entropy weight of the i1th index is: Weight vector W (1) =(X1,X2,...X i ,...,X m ); After repeating the above steps r times, the average of the r entropy weights of each indicator is the final objective weight;

[0040] S33, for the weight vector set {W (1) ,W (2) ,...,W (D)}D weight vectors are combined linearly to obtain a feasible weight vector: Among them, F is a feasible weight vector, α d is the weight coefficient, is the transpose of the d-th weight vector; according to game theory, the most suitable combination weight F* can be found in the feasible weight vector set.

[0041] Furthermore, in step S32, the method for normalizing the eigenvalue matrix to obtain the relative superiority matrix is:

[0042] When the evaluation index is positively correlated with the overall goal, the benefit-based normalization formula is used:

[0043]

[0044] in, is the normalized index value; are the evaluation index values ​​of different evaluation objects under the same evaluation index. The maximum and minimum values ​​in ;

[0045] When the evaluation index is negatively correlated with the overall goal, the cost normalization formula is used:

[0046]

[0047] Furthermore, in step S33, the method for finding the most appropriate combined weight F* in the feasible weight vector set according to game theory is:

[0048] Convert the original problem into weight coefficient α d Make the feasible weight vector F and each The sum of the gaps is the smallest, that is Where d = 1, 2, ..., D; solving it yields the first-order derivative condition:

[0049]

[0050] The corresponding linear equations are:

[0051]

[0052] For (α1,α2,...,α D ) Calculate the final coefficients: The combined weight is:

[0053]

[0054] Furthermore, step S4 is specifically as follows:

[0055] S41. Collect relevant information on the modernization of embankment management, analyze the level of achievement of each qualitative indicator, and invite experts to evaluate and score each qualitative indicator; collect quantitative basic indicator evaluation data and evaluate the management unit;

[0056] S42, converting the evaluation value of the comment level of each qualitative indicator into a cloud model according to step S2; using the reverse cloud algorithm to calculate the t evaluation values ​​of each quantitative indicator in different states to obtain a cloud model;

[0057] S43. Generate a random cloud droplet for the qualitative indicator cloud model included in the secondary indicator; replace the qualitative comments with the score of this cloud droplet, and form a quantitative evaluation matrix with the original quantitative indicator data; perform a weighted calculation on the replaced quantitative evaluation matrix and the indicator combination weight obtained in step S3 to obtain a random value for the secondary indicator;

[0058] S44, repeat step S43 nt times to obtain nt random values ​​of the secondary indicators, and perform reverse cloud computing on them to obtain the secondary indicator evaluation cloud model;

[0059] S45. Using the floating cloud algorithm of the cloud model, the low-level indicator evaluation cloud model is calculated to obtain a high-level indicator evaluation cloud model, thereby promoting the low-level concepts between adjacent indicator levels to high-level concepts; after obtaining all the second-level indicator evaluation cloud models, floating cloud computing is performed twice in iteration to obtain the first-level indicator cloud models and the overall target cloud model;

[0060] S46. Use the division of the cloud model to calculate the overall goal cloud model and the first-level indicator cloud model with the corresponding target values, and obtain the overall goal realization cloud and the first-level indicator realization cloud; compare each realization cloud with the scale cloud of the evaluation grade standard to determine the overall goal modernization realization degree evaluation grade and the first-level indicator modernization realization degree evaluation grade.

[0061] The present invention provides a levee engineering modernization evaluation system based on a combination of cloud model and entropy weight, comprising:

[0062] Index system construction module, used to build a modern evaluation index system for embankment project management;

[0063] The conversion model building module is used to establish the evaluation index data and cloud model conversion model based on the modernization evaluation index system of embankment engineering management;

[0064] The indicator weight determination module is used to determine the weight of each indicator based on the combined weighting of the cloud model, entropy weight method and game theory;

[0065] Evaluation model construction module, used to build a modern evaluation model for embankment project management based on cloud theory;

[0066] The evaluation module is used to conduct specific evaluation using the embankment project modernization evaluation model based on the combination of cloud model and entropy weight.

[0067] A device of the present invention includes a memory and a processor, wherein:

[0068] a memory for storing computer programs capable of running on the processor;

[0069] The processor is used to execute the steps of the above-mentioned method for evaluating the modernization of embankment projects based on a combination of cloud model and entropy weight when running the computer program.

[0070] Beneficial effects: Compared with the existing technology, the significant technical effects of the present invention are: conducting research on the modernization of embankment engineering in terms of standardized management of embankment engineering, modernization of embankment engineering facilities and equipment management, information management of embankment engineering, embankment scheduling and operation, emergency response capabilities and water ecological management; adopting a comprehensive evaluation method based on a combination of cloud model and entropy weight to evaluate the scheme; using game theory to calculate the combined weight to ensure that the total difference between the combined weight and the original weight is minimized, making the evaluation result more scientific and reasonable; the method of the present invention can be better applied in real life. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 Flow chart of the method of the present invention;

[0072] Figure 2 It is a conversion diagram between qualitative indicator rating comments and cloud model;

[0073] Figure 3 This is a comparison chart of the overall goal and the degree of modernization of each first-level indicator and the scale cloud. DETAILED DESCRIPTION

[0074] The technical solution of the present invention is further described in detail below with reference to the accompanying drawings:

[0075] The modern evaluation method of levee engineering based on the combination of cloud model and entropy weight in this invention establishes an evaluation index system mainly for levees above the second level, such as Figure 1 The specific steps are as follows:

[0076] S1. Establish a modern evaluation index system for embankment project management; specifically including:

[0077] S11. Using the frequency analysis method, based on the content and spirit of the "Water Conservancy Project Management Assessment Method", relevant documents on provincial river embankments, and relevant literature research, the indicators listed in existing relevant documents and literature research are clustered and summarized, and forward-looking expansion is carried out;

[0078] S12. Use theoretical analysis, guided by the connotation and content of the evaluation object, starting from comprehensiveness, necessity, and systematicness, combining experience and deficiencies in management practice, and selecting representative indicators based on the components of the evaluation object and their internal connections;

[0079] S13. The Delphi method was used to consult experts and the principal component analysis method was used to optimize the indicators and construct an evaluation index system for embankment engineering management modernization, which includes five first-level indicators, 22 second-level indicators, and 99 third-level indicators.

[0080] S14. Based on the establishment of a modernized evaluation index system for embankment project management, determine the corresponding relationship between the grade comments and value ranges for all qualitative indicators, except for the five quantitative indicators of a 50% cleaning rate within the demarcation area, a 90% cleaning rate within the title confirmation area, a 50% greening coverage rate within the demarcation area, a 90% greening coverage rate within the title confirmation area, and a 90% soil and water loss control rate, as shown in Table 1:

[0081] Table 1 Index grade comments and their corresponding score ranges

[0082]

[0083] S2. Based on the constructed modern evaluation index system for embankment engineering management, the evaluation index data and cloud model are converted. Specifically, the conversion includes:

[0084] S21. The cloud model uses three independent parameters: expectation Ex, entropy En, and hyperentropy H. e To express a qualitative concept. Indicator data is divided into quantitative indicator data and qualitative indicator data. The t evaluation values ​​of a quantitative indicator in different states can be integrated into a cloud model to express it. The expected Ex, entropy En, and super entropy H can be obtained by using the reverse cloud generator. e ;

[0085]

[0086]

[0087]

[0088]

[0089] Among them, Ex is the expectation of the cloud model; En is the entropy of the cloud model; H e is the super entropy of the cloud model; x i is the i-th evaluation value of the quantitative indicator; S 2 is the variance of the quantitative indicator; t is the number of evaluation values ​​of the quantitative indicator;

[0090] It can also be obtained by simplifying the calculation through formula (5), in which case the quantitative index corresponds to the super entropy H of the cloud model: e is a specified constant;

[0091]

[0092] S22. Qualitative indicators can be set with different numbers and actual meanings of evaluation levels as needed. The domain corresponding to the evaluation level of qualitative indicators is Ω, and Ω′ is a division of the domain Ω:

[0093]

[0094] Among them, Ω i is the domain corresponding to the i-th level review level; k is the number of review levels.

[0095] The normal evaluation cloud model of the i-th level qualitative comment grade on the partition domain is called the grade evaluation cloud model, and its digital characteristics are T i =(Ex i ,En i ,He i ); T1 and T k When they are defined as half falling cloud and half rising cloud respectively, the evaluation value interval of level i is [a i ,b i ], where i = 1, 2, ..., k, based on the 3En rule of the normal cloud model, the cloud parameter of the i-th level (Ex i ,Eni ) According to formula (7), the super entropy He i is a constant;

[0096]

[0097] According to the corresponding relationship between the grade comments and value intervals of the qualitative basic indicators in Table 1, the qualitative indicator data and the cloud model are converted according to formula (7), and the super entropy He is taken. i As a constant of 0.005, the cloud model parameters of the review level are obtained, as shown in Table 2. The generated cloud map is as follows Figure 2 shown.

[0098] Table 2 Cloud model parameters for qualitative indicator rating comments

[0099]

[0100] When an indicator of the evaluation object has multiple grade comment values, the expected value Ex, entropy En, and super entropy H of the grade evaluation cloud corresponding to the indicator are calculated using formula (8). e ;

[0101]

[0102] Where H is the number of grade evaluation values; Ex i′ is the expectation of the grade evaluation cloud corresponding to the i′th grade comment value; En i′ is the entropy of the grade evaluation cloud corresponding to the i′th grade comment value; He i′ is the super entropy of the grade evaluation cloud corresponding to the i′th grade comment value.

[0103] S3. Determine the combined weights of various indicators based on the cloud model, entropy weight method, and game theory. Specifically,

[0104] S31. Invite n experts to give the weight distribution results of m indicators: Where i1 = 1, 2, ..., m, j = 1, 2, ..., n; using the reverse cloud generator that does not require the degree of certainty μ(x), the weighted cloud model parameters are obtained by calculating the n weighted evaluation values ​​of the indicator i1 The weighted cloud model of index i1 is tested for accuracy and its parameters are corrected. The forward cloud algorithm is used to randomly generate q cloud droplets for the corrected weighted cloud model of index i1, and the score corresponding to the maximum certainty of the cloud droplet point is used as the weight calculation value ω of index i1. i ; Normalize the weight calculation values ​​of m indicators to obtain the weight set:

[0105] S32. For each qualitative rating of the indicator, according to Table 1 and Figure 2Obtain cloud model parameters; use the forward cloud algorithm to generate a random cloud droplet for each cloud model and replace it with the qualitative comments in the evaluation result matrix; when there are w groups of original data for m evaluation indicators, the evaluation indicator eigenvalue matrix is: Where j1 = 1, 2, ..., w; using formula (9) and (10) to normalize the eigenvalue matrix, we can get the relative superiority matrix R = (r i1j1 ) m×w ,in is the normalized value of the indicator; calculate the entropy value The entropy weight of the i1th indicator is: Weight vector W (1) =(X1,X2,...X i ,...,X m ); After repeating the above steps r times, the average of the r entropy weights of each indicator is the final objective weight.

[0106] When the evaluation index is positively correlated with the overall goal, the benefit-based normalization formula is used:

[0107]

[0108] in, is the normalized index value; are the evaluation index values ​​of different evaluation objects under the same evaluation index. The maximum and minimum values ​​in ;

[0109] When the evaluation index is negatively correlated with the overall goal, the cost normalization formula is used:

[0110]

[0111] S33, for the weight vector set {W (1) ,W (2) ,...,W (D)}D weight vectors are combined linearly to obtain a feasible weight vector: Among them, F is a feasible weight vector, α d is the weight coefficient, is the transpose of the d-th weight vector; according to game theory, the most suitable combination weight F* can be found in the feasible weight vector set. Its main implementation method is to transform the original problem into the weight coefficient α d Make F and each The sum of the gaps is the smallest, that is Where d = 1, 2, ..., D; solving it yields the first-order derivative condition:

[0112]

[0113] The corresponding linear equations are:

[0114]

[0115] For (α1,α2,...,α D ) Calculate the final coefficients: The combined weight is:

[0116]

[0117] S4. Construct a modern evaluation model for embankment engineering management based on cloud theory; specifically including:

[0118] S41. Comprehensively collect relevant information on the modernization of unit management, analyze the level of each qualitative basic indicator, and invite experts to evaluate and score each qualitative indicator. Collect quantitative basic indicator evaluation data and evaluate the management unit;

[0119] S42, converting the evaluation value of the comment level of each qualitative indicator into a cloud model according to step S2; using the reverse cloud algorithm to calculate the t evaluation values ​​of each quantitative indicator in different states to obtain a cloud model;

[0120] S43. Generate a random cloud droplet for the qualitative indicator cloud model included in the secondary indicator; replace the qualitative comments with the score of this cloud droplet, and form a quantitative evaluation matrix with the original quantitative indicator data; perform a weighted calculation on the replaced quantitative evaluation matrix and the indicator combination weight obtained in step S3 to obtain a random value for the secondary indicator;

[0121] S44, repeat step S43 for nt times to obtain nt random values ​​of the secondary indicators, and perform reverse cloud computing on them to obtain the secondary indicator evaluation cloud model;

[0122] S45. Use the floating cloud algorithm of the cloud model to calculate the low-level indicator evaluation cloud model to obtain the high-level indicator evaluation cloud model, so as to realize the promotion of low-level concepts to high-level concepts between adjacent indicator levels; after obtaining all the secondary indicator evaluation cloud models, perform floating cloud computing twice in iteration to obtain the primary indicator cloud models and the overall target cloud model.

[0123] S46. Use the division method of the cloud model to calculate the overall target cloud model and the first-level indicator cloud with the corresponding target values, and obtain the overall target realization cloud and the first-level indicator realization cloud; compare each realization cloud with the scale cloud of the evaluation grade standard to determine the overall target modernization realization degree evaluation grade and the first-level indicator modernization realization degree evaluation grade.

[0124] S5. Use the cloud theory-based modern evaluation model for embankment project management to conduct specific evaluation;

[0125] Based on the current status of levee engineering management within the study area, the overall evaluation objectives for levee management modernization and the target values ​​for each primary indicator were determined. This model was used to comprehensively evaluate the modernization level of levee management units within the study area. Based on this, a cloud model for the levee management modernization grade evaluation standard was determined.

[0126] The present invention provides a levee engineering modernization evaluation system based on a combination of cloud model and entropy weight, comprising:

[0127] Index system construction module, used to build a modern evaluation index system for embankment project management;

[0128] The conversion model building module is used to establish the evaluation index data and cloud model conversion model based on the modernization evaluation index system of embankment engineering management;

[0129] The indicator weight determination module is used to determine the weight of each indicator based on the combined weighting of the cloud model, entropy weight method and game theory;

[0130] Evaluation model construction module, used to build a modern evaluation model for embankment project management based on cloud theory;

[0131] The evaluation module is used to conduct specific evaluation using the embankment project modernization evaluation model based on the combination of cloud model and entropy weight.

[0132] A device of the present invention includes a memory and a processor, wherein:

[0133] a memory for storing computer programs capable of running on the processor;

[0134] The processor is used to execute the steps of the above-mentioned method for evaluating the modernization of embankment projects based on a combination of cloud model and entropy weight when running the computer program, and achieve technical effects consistent with the above-mentioned method.

[0135] A storage medium of the present invention stores a computer program, which, when executed by at least one processor, implements the steps of the above-mentioned method for evaluating the modernization of embankment projects based on a combination of cloud models and entropy weights, and achieves technical effects consistent with the above-mentioned method.

[0136] Example:

[0137] According to the above step S1:

[0138] The contents of the modernization of embankment project management include modernization of management concepts, modernization of management systems and mechanisms, institutionalization, standardization and legalization of management, standardization of engineering facilities and management facilities, modernization of management methods, modernization of management teams, and water culture construction.

[0139] This index system primarily focuses on dike management units at or above the second level, and consists of a target layer (A), a criterion layer (B), and an indicator layer (C and D). The target layer (A) represents the overall assessment objective; the criterion layer (B) comprises sub-objectives, including five first-level indicators (B1-B5); and the indicator layer (C and D) represents a more detailed expression of the overall objective in various aspects, comprising 22 second-level indicators (C1-C22), 99 third-level indicators, and key assessment items (D1-D99).

[0140] Basic indicators can be divided into two categories: qualitative and quantitative evaluation. Quantitative evaluation includes the five indicators D90, D91, D93, D94 and D96 in B5, and the other indicators are qualitative evaluation.

[0141] Table 2 Standardized management of embankment projects (B1)

[0142]

[0143]

[0144] Table 3 Modernization of embankment engineering facilities and equipment management (B2)

[0145]

[0146]

[0147] Table 4 Information management of embankment projects (B3)

[0148]

[0149] Table 5 Dike Operation and Emergency Response Capacity (B4)

[0150]

[0151]

[0152] Table 6 Water Ecological Management (B5)

[0153]

[0154] According to the above step S2:

[0155] According to the evaluation index and cloud model conversion rules, the evaluation results are converted into a cloud model. According to the cloud model calculation method of the ascending concept level, the secondary indicator cloud model is obtained as shown in Table 7.

[0156] Table 7 Statistics of cloud parameters of secondary indicators of W levee

[0157]

[0158]

[0159] According to the above step S3:

[0160] Ten experts were invited to score the relative weights of each indicator layer in the indicator system, and the subjective weight calculation method based on the cloud model was used for calculation. The results are shown in Table 8.

[0161] Table 8 Subjective weight calculation results

[0162]

[0163]

[0164]

[0165] The subjective weights of basic indicators relative to the overall target are obtained through conversion, as shown in Table 9.

[0166] Table 9A-D Subjective Weight Statistics

[0167]

[0168] In the province where the levee management unit (W Levee) in the case analysis is located, 11 representative and typical first-level levee management units and 21 second-level levee management units were selected for evaluation. Qualitative indicator evaluation scores and quantitative indicator data were obtained. According to the objective weight determination method of the coupled cloud model and entropy weight method, the objective weight of the basic indicators relative to the overall target was calculated 1000 times, as shown in Table 10.

[0169] Table 10A-D Objective Weight Statistics

[0170]

[0171] The consistency test of the two weights of the basic indicators is carried out, d(W (1) ,W (2) ) = 0.0459, indicating good weight consistency. The combined weight calculation method based on game theory was applied to obtain the combined weights of the basic indicators relative to the overall goal. The relative weights of the indicators at each level were then calculated based on the indicator system structure, as shown in Table 11.

[0172] Table 11 Combination weight statistics

[0173]

[0174]

[0175]

[0176] According to the above step S4:

[0177] The floating cloud algorithm of formula (14) is used to calculate the cloud model of the secondary indicator to obtain the cloud model parameters of each primary indicator. The statistical table is shown in Table 12.

[0178]

[0179] in, is the cloud parameter of the i1th base cloud, is the weight of the concept corresponding to the i1th base cloud, and (Ex, En, He) are the parameters of the comprehensive concept corresponding cloud model.

[0180] Table 12 Statistics of cloud model parameters for the first-level index of the W levee

[0181]

[0182]

[0183] Similarly, the overall target cloud model parameters of W Dam are (0.830, 0.0171, 0.005).

[0184] The total target cloud and each first-level indicator cloud are divided by their target values ​​using the division method of formula (15), thereby obtaining the cloud model of their management modernization realization degree, as shown in Table 12.

[0185]

[0186] Table 12 Cloud parameter statistics of overall goals and primary indicators

[0187]

[0188] Calculate the similarity between the first-level indicator achievement degree cloud and the overall goal achievement degree cloud and the level cloud in the scale cloud, as shown in Table 13. And generate the cloud map as shown in the attached Figure 3 As shown, (a)-(e) are the comparison charts of the B1-B5 first-level indicator realization cloud and the ruler cloud respectively, and (f) is the comparison chart of the overall goal realization cloud and the ruler cloud.

[0189] Table 13 Comparison of the overall goal and primary indicator achievement level cloud and scale cloud

[0190]

[0191]

[0192] Based on Table 13, the evaluation level of each indicator is determined based on the highest similarity between each indicator and the respective level cloud in the scale cloud. The overall goal of achieving modernization management has reached the "preliminary achievement" level. The achievement levels of each first-level indicator are: (basic achievement, preliminary achievement, unachieved, basic achievement, unachieved), as shown in Table 14.

[0193] Table 14 Determination of overall goals and levels of each first-level indicator

[0194]

[0195] In summary, the present invention's method for evaluating levee project modernization based on a combination of cloud models and entropy weights includes: constructing an evaluation index system for levee project management modernization; establishing a conversion model between evaluation index data and cloud models based on the constructed levee project management modernization evaluation index system; determining the weights of various indicators through combined weighting based on the cloud model, entropy weighting, and game theory; constructing a levee project management modernization evaluation model based on cloud theory; and conducting specific evaluations using the levee project management modernization evaluation model based on cloud theory. This invention clarifies the connotation and content of levee project management modernization, constructs a multi-level evaluation index system, and establishes an evaluation model based on a combination of cloud models and entropy weights, providing a more comprehensive and scientific evaluation model for levee management modernization.

Claims

1. A method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight, characterized in that: The following steps are involved: S1. Construct a modern evaluation index system for embankment project management; S2. According to the constructed embankment engineering management modernization evaluation index system, the evaluation index data is converted into the cloud model; wherein, the cloud model uses three independent parameters: expectation Ex, entropy En, and super entropy H. e To express a qualitative concept; evaluation index data includes quantitative index data and qualitative index data. T evaluation values ​​of a quantitative index in different states are integrated into a cloud model to represent it; qualitative indicators are set with comment levels according to needs. According to the correspondence between the level comments and value ranges of qualitative indicators, the qualitative indicator data and the cloud model are converted; S3. Based on the combined weighting of cloud model, entropy weight method and game theory, the combined weight of each indicator is determined; first, the weight distribution result of each indicator is determined based on expert theory, the weight cloud model parameters of each indicator are calculated, and the weight cloud model is subjected to accuracy test and parameter correction. The weight calculation value of each indicator in the corrected weight cloud model is calculated, and the weight calculation values ​​of all indicators are normalized to obtain a weight set; secondly, the cloud model of each indicator is obtained according to the method of step S2, and the entropy weight of each indicator is calculated to form a weight vector, and the final objective weight is obtained; specifically, for each qualitative grade comment of the indicator, the cloud model parameters are obtained according to the transformation of step S2; a random cloud droplet is generated for each cloud model using the forward cloud algorithm, and it replaces the qualitative comment in the evaluation result matrix; when there are w groups of original data for m evaluation indicators, there is an evaluation indicator eigenvalue matrix: Where j1=1,2,...,w; normalize the eigenvalue matrix to get the relative superiority matrix in is the normalized value of the indicator; calculate the entropy value is the normalized index value; the entropy weight of the i1th index is: Weight vector W (1) =(X1,X2,...X i ,...,X m ); After repeating the above steps r times, the average of the r entropy weights of each indicator is the final objective weight; finally, any linear combination of multiple weight vectors in the weight vector set is performed to obtain a feasible weight vector, and the most appropriate combination weight is found in the feasible vector set according to game theory; S4. Construct a cloud theory-based evaluation model for the modernization of levee project management. First, collect relevant information and obtain cloud models for each indicator according to the method in step S2. Second, replace the qualitative comments with the scores of the qualitative indicator cloud model, form a quantitative evaluation matrix with the original quantitative indicator data, and perform weighted calculation with the indicator combination weights obtained in step S3. Repeat the operation nt times to obtain a secondary indicator evaluation cloud model. Finally, calculate the secondary indicator evaluation cloud model to obtain the overall goal realization cloud model and the primary indicator realization cloud model, and compare them with the scale cloud model of the evaluation grade standard to determine the evaluation grade of the overall goal modernization realization degree and the evaluation grade of the primary indicator modernization realization degree. S5. Use the cloud theory-based modern evaluation model for embankment engineering management to conduct specific evaluation.

2. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: Step S1 is specifically as follows: S11. Use frequency analysis to cluster and summarize the indicators listed in existing relevant documents and literature research, and conduct forward-looking expansion; S12. Use theoretical analysis to select representative indicators based on the components of the evaluation object and their internal connections; S13. The Delphi method was used to consult experts and the principal component analysis method was used to optimize the indicators and construct an evaluation index system for embankment engineering management modernization, which includes five first-level indicators, 22 second-level indicators, and 99 third-level indicators. S14. Based on the constructed embankment engineering management modernization evaluation index system, determine the corresponding relationship between the grade comments and value ranges of all qualitative indicators except the five quantitative indicators: the cleaning rate within the demarcation area reaches 50%, the cleaning rate within the title confirmation area reaches 90%, the greening coverage rate within the demarcation area reaches 50%, the greening coverage rate within the title confirmation area reaches 90%, and the soil and water loss control rate reaches 90%, as shown in Table 1: Table 1 Index grade comments and their corresponding score ranges 3. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: The method for establishing the quantitative indicator data and cloud model conversion model in step S2 is: The t evaluation values ​​of a quantitative indicator in different states can be integrated into a cloud model to represent it, and the expected Ex, entropy En, and super entropy H are obtained by using the reverse cloud generator. e The calculation formula is: Among them, Ex is the expectation of the cloud model; En is the entropy of the cloud model; H e is the super entropy of the cloud model; x i is the i-th evaluation value of the quantitative indicator; S 2 is the variance of the quantitative indicator; t is the number of evaluation values ​​of the quantitative indicator; Or it can be obtained by simplifying the calculation using the following formula. In this case, the quantitative index corresponds to the super entropy H of the cloud model: e is a specified constant; 4. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: The method for establishing the qualitative indicator data and cloud model conversion model in step S2 is: Qualitative indicators are set with different numbers and actual meanings of evaluation levels according to needs. The domain corresponding to the evaluation level of qualitative indicators is Ω, and Ω′ is a division of Ω: Among them, Ω i is the domain corresponding to the i-th level of comment; k is the number of comment levels; The normal evaluation cloud model of the i-th level qualitative comment grade on the partition domain is called the grade evaluation cloud model, and its digital feature is T i =(Ex i ,En i ,He i ); T1 and T k When they are defined as half falling cloud and half rising cloud respectively, the evaluation value interval of level i is [a i ,b i ], where i = 1, 2, ..., k, based on the 3En rule of the normal cloud model, the cloud parameter of the i-th level (Ex i ,En i ) According to the following formula, the super entropy He i is a constant; According to the corresponding relationship between the grade comments and value ranges of the qualitative basic indicators, the qualitative indicator data and the cloud model are converted according to the above formula to obtain the cloud model parameters of the comment grade; When an indicator of the evaluation object has multiple grade comment values, the following formula is used to calculate the expected Ex, entropy En, and super entropy H of the grade evaluation cloud corresponding to the indicator. e ; Where H is the number of grade evaluation values; Ex i′ is the expectation of the grade evaluation cloud corresponding to the i′th grade comment value; En i′ is the entropy of the grade evaluation cloud corresponding to the i′th grade comment value; He i′ is the super entropy of the grade evaluation cloud corresponding to the i′th grade comment value.

5. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: The specific method for calculating the weight set in step S3 is: Invite n experts to give the weight distribution results of m indicators: Where i1 = 1, 2, ..., m, j = 1, 2, ..., n; using the reverse cloud generator that does not require the degree of certainty μ(x), the weighted cloud model parameters are obtained by calculating the n weighted evaluation values ​​of the indicator i1 The weighted cloud model of index i1 is tested for accuracy and its parameters are corrected. The forward cloud algorithm is used to randomly generate q cloud droplets for the corrected weighted cloud model of index i1, and the score corresponding to the maximum certainty of the cloud droplet point is used as the weight calculation value ω of index i1. i ; Normalize the weight calculation values ​​of m indicators to obtain the weight set: The combination weight calculation method is: for the weight vector set {W (1) ,W (2) ,...,W (D) }D weight vectors are combined linearly to obtain a feasible weight vector: Among them, F is a feasible weight vector, α d is the weight coefficient, is the transpose of the d-th weight vector; according to game theory, the most suitable combination weight F* can be found in the feasible weight vector set.

6. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: The method for normalizing the eigenvalue matrix in step S3 to obtain the relative superiority matrix is: When the evaluation index is positively correlated with the overall goal, the benefit-based normalization formula is used: in, is the normalized index value; are the evaluation index values ​​of different evaluation objects under the same evaluation index. The maximum and minimum values ​​in ; When the evaluation index is negatively correlated with the overall goal, the cost normalization formula is used:

7. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 5 is characterized in that: According to game theory, the method to find the most appropriate combination weight F* in the feasible weight vector set is: Convert the original problem into weight coefficient α d Make the feasible weight vector F and each The sum of the gaps is the smallest, that is Where d = 1, 2, ..., D; solving it yields the first-order derivative condition: The corresponding linear equations are: For (α1,α2,...,α D ) Calculate the final coefficients: The combined weight is:

8. The method for evaluating embankment engineering modernization based on a combination of cloud model and entropy weight according to claim 1 is characterized in that: Step S4 is specifically as follows: S41. Collect relevant information on the modernization of embankment management, analyze the level of achievement of each qualitative basic indicator, and invite experts to evaluate and score each qualitative indicator; collect evaluation data on quantitative basic indicators and evaluate the management units; S42, converting the evaluation value of the comment level of each qualitative indicator into a cloud model according to step S2; using the reverse cloud algorithm to calculate the t evaluation values ​​of each quantitative indicator in different states to obtain a cloud model; S43. Generate a random cloud droplet for the qualitative indicator cloud model included in the secondary indicator; replace the qualitative comments with the score of this cloud droplet, and form a quantitative evaluation matrix with the original quantitative indicator data; perform a weighted calculation on the replaced quantitative evaluation matrix and the indicator combination weight obtained in step S3 to obtain a random value for the secondary indicator; S44, repeat step S43 nt times to obtain nt random values ​​of the secondary indicators, and perform reverse cloud computing on them to obtain the secondary indicator evaluation cloud model; S45. Using the floating cloud algorithm of the cloud model, the low-level indicator evaluation cloud model is calculated to obtain a high-level indicator evaluation cloud model, thereby promoting the low-level concepts between adjacent indicator levels to high-level concepts; After obtaining all the secondary indicator evaluation cloud models, floating cloud computing is performed twice in iteration to obtain the primary indicator cloud models and the overall target cloud model; S46. Use the division of the cloud model to calculate the overall goal cloud model and the first-level indicator cloud model with the corresponding target values, and obtain the overall goal realization cloud and the first-level indicator realization cloud; compare each realization cloud with the scale cloud of the evaluation grade standard to determine the overall goal modernization realization degree evaluation grade and the first-level indicator modernization realization degree evaluation grade.

9. A levee engineering modernization evaluation system based on a combination of cloud model and entropy weight, characterized in that: include: Index system construction module, used to build a modern evaluation index system for embankment project management; The conversion model construction module is used to convert the evaluation index data and the cloud model according to the constructed embankment engineering management modernization evaluation index system; the cloud model uses three independent parameters: expectation Ex, entropy En, and super entropy H. e To express a qualitative concept; evaluation index data includes quantitative index data and qualitative index data. T evaluation values ​​of a quantitative index in different states are integrated into a cloud model to represent it; qualitative indicators are set with comment levels according to needs. According to the correspondence between the level comments and value ranges of qualitative indicators, the qualitative indicator data and the cloud model are converted; The indicator weight determination module is used to determine the combined weights of various indicators based on the combined weighting of the cloud model, entropy weight method and game theory; first, the weight distribution results of each indicator are determined based on expert theory, the weight cloud model parameters of each indicator are calculated, and the weight cloud model is subjected to accuracy test and parameter correction, and the weight calculation value of each indicator in the corrected weight cloud model is calculated, and the weight calculation values ​​of all indicators are normalized to obtain a weight set; secondly, the cloud model of each indicator is obtained according to the conversion model construction module, and the entropy weight of each indicator is calculated to form a weight vector, and the final objective weight is obtained; specifically, for each qualitative grade comment of the indicator, the cloud model parameters are obtained according to the conversion model construction module; a random cloud droplet is generated for each cloud model using the forward cloud algorithm, and it replaces the qualitative comment in the evaluation result matrix; when there are w groups of original data for m evaluation indicators, there is an evaluation indicator eigenvalue matrix: X=(x i1j1 ) m×w , where j1=1,2,...,w; normalize the eigenvalue matrix to obtain the relative superiority matrix in is the normalized value of the indicator; calculate the entropy value is the normalized index value; the entropy weight of the i1th index is: Weight vector W (1) =(X1,X2,...X i ,...,X m ); After repeating the above steps r times, the average of the r entropy weights of each indicator is the final objective weight; finally, any linear combination of multiple weight vectors in the weight vector set is performed to obtain a feasible weight vector, and the most appropriate combination weight is found in the feasible vector set according to game theory; The evaluation model construction module is used to construct an evaluation model for the modernization of levee project management based on cloud theory. First, relevant information is collected and the cloud model of each indicator is obtained according to the conversion model construction module. Second, the qualitative comments are replaced by the scores of the qualitative indicator cloud model, and a quantitative evaluation matrix is ​​formed with the original quantitative indicator data. The weighted calculation is performed with the indicator combination weight obtained by the indicator weight determination module, and the operation is repeated nt times to obtain a secondary indicator evaluation cloud model. Finally, the secondary indicator evaluation cloud model is calculated to obtain the overall goal realization cloud model and the first-level indicator realization cloud model, and compared with the scale cloud model of the evaluation grade standard to determine the overall goal modernization realization degree evaluation grade and the first-level indicator modernization realization degree evaluation grade. The evaluation module is used to use the cloud theory-based levee project management modernization evaluation model for specific evaluation.

10. A device, characterized in that: comprising a memory and a processor, wherein: a memory for storing computer programs capable of running on the processor; A processor is used to execute the steps of a levee engineering modernization evaluation method based on a combination of cloud model and entropy weight as described in any one of claims 1 to 8 when running the computer program.

Images