A clustering method for safety assessment of earth-rock dams based on sub-items and sub-positions

Abstract

The invention discloses a clustering safety evaluation method for earth-rock dams based on sub-items and sub-parts, comprising the following steps: step 1: constructing an earth-rock dam safety evaluation index system based on sub-items and sub-parts; step 2: determining the weights of subjective and objective indicators for each item and each part; step 3: adjusting the weights according to a game theory method to obtain an adjusted combined weight; step 4: determining the health level of historical earth-rock dams; step 5: calculating the membership of each indicator according to monitoring data; step 6: obtaining an evaluation coefficient according to the membership, and obtaining an evaluation result by adopting cluster analysis in combination with the health level. In view of the problem that the current dam health evaluation conclusions are not very practical, the invention refines the earth-rock dam into sub-items according to different structural and damage characteristics, adopts game theory to dynamically adjust and modify the subjective and objective weights, adopts a clustering method to analyze the result characteristics between years, and provides safety evaluation results for different parts and projects.

Description

Technical Field

[0001] The present invention relates to the technical field of dam safety, and in particular to an earth-rock dam evaluation method based on clustering of sub-items and sub-positions. Background Art

[0002] Earth-rockfill dam safety assessment is a comprehensive evaluation of the dam's operational performance through research into its structure and operation. This evaluation identifies key issues and identifies key areas of concern and weaknesses. Currently, the main methods include expert evaluation and risk-based comprehensive assessment.

[0003] Comprehensive dam evaluations conducted by experts are currently the most common and reliable method for dam safety assessments. Regular dam safety inspections every five years are essential. However, this method is labor-intensive and time-consuming, making it difficult to achieve real-time, dynamic, and process-based assessments of dam safety, and presents potential safety risks.

[0004] Risk analysis considers the possibility of a dam failure and the cumulative consequences. It involves five steps: risk identification, risk estimation, risk assessment, risk transfer, and risk decision-making. Risk identification involves proposing possible dam failure paths through in-depth analysis of the dam site's geology and dam structure. Risk estimation involves calculating the probability of a dam failure and the resulting economic, life, social, and environmental damages. Risk assessment involves evaluating the cumulative risk and losses of a dam failure. Risk transfer involves mitigating the risk through the transfer of consequences. Risk decision-making involves iterating these steps to propose a number of feasible risk management options. In recent years, dam risk analysis methods have been widely studied and applied within the dam engineering community. However, these methods, which primarily consider the potential for dam failure and the resulting risks and losses, offer clear concepts and strong guidance. However, the calculations involve both quantitative and non-quantitative components, and initial model development is complex, requiring specialized simulations. These methods are difficult for general engineers to master and difficult to implement in real time.

[0005] The comprehensive evaluation method follows the principle of determining an evaluation system based on the evaluation objectives and the feasibility of quantitatively implementing indicators; determining indicator weights and membership functions; and employing calculation methods to evaluate the objectives. Due to its clear concept and well-defined indicator system, this method has been widely applied in numerous fields, including water conservancy, civil engineering, mining, and the environment. However, it suffers from the following issues: the specification and weighting of indicators in the evaluation method are highly subjective; the boundaries of deformation indicators during earth-rockfill dam construction are difficult to define; and the evaluation conclusions are presented as scores, followed by safety grading based on these boundaries, which is of limited practical value for project guidance. Summary of the Invention

[0006] Aiming at the problems existing in the prior art, the present invention provides an earth-rock dam evaluation method based on clustering of sub-items and sub-positions.

[0007] The technical solution adopted by the present invention is: a clustering earth-rock dam evaluation method based on sub-items and sub-positions, comprising the following steps:

[0008] Step 1: Construct a safety evaluation index system for earth-rockfill dams by item and section;

[0009] Step 2: Determine the weights of subjective and objective indicators for each item and department; determine the objective indicator weight ω2 based on monitoring data, and determine the subjective indicator weight ω1 based on preset expert experience;

[0010] Step 3: Adjust the weights according to the game theory method to obtain the adjusted combined weight ω;

[0011] Step 4: Determine the health level of the historical earth-rockfill dam;

[0012] Step 5: Calculate the membership degree of each indicator based on the monitoring data;

[0013] Step 6: Obtain the evaluation coefficient based on the membership degree obtained in step 5, and use cluster analysis to obtain the evaluation results in combination with the health level in step 4.

[0014] Furthermore, the safety index evaluation system for layered earth-rock dams is as follows:

[0015] The safety evaluation set of earth-rock dam is denoted as C = {C1, C2, ..., C n}, C i is the project layer, n is the number of projects; C i ={c i1 , c i2 ,…,c im}, where c ij is the jth evaluation index of the i-th layer, and m is the number of indicators;

[0016] The project layer includes flood resistance, structural stability, seepage stability, seismic safety and metal structure; flood resistance includes dam crest height review and downstream flood review; structural stability includes dam slope stability, dam body deformation, anti-seepage body deformation, other structural deformation, contact deformation, and bank slope near the dam; seepage stability includes seepage pressure, seepage gradient, seepage volume and seepage water quality; seismic safety includes dam seismic safety factor and soil liquefaction possibility; metal structure includes the working condition of steel gate and hoist.

[0017] Furthermore, the calculation method of the combination weight is as follows:

[0018] ω=δ k ω1+(1-δ k )ω2

[0019] Among them, δ k is the weight distribution coefficient of objective weight and subjective weight, δ kDetermined based on game theory methods.

[0020] Furthermore, the subjective indicator weight ω1 and the objective indicator weight ω2 are determined according to the entropy weight method.

[0021] Furthermore, the evaluation coefficient calculation process is as follows:

[0022] Calculation project layer C i Membership degree x i :

[0023]

[0024] Where: x ij is the indicator c ij The corresponding membership degree, ω ij is the indicator c ij The corresponding weight;

[0025] Evaluation coefficient T1 is

[0026]

[0027] Where: δT is the adjustment coefficient, which is determined based on the safety evaluation index of earth-rock dam.

[0028] Furthermore, in step 6, the inter-annual correlation of the earth-rock dam safety evaluation index is obtained according to the clustering method, and the inter-annual correlation is combined with the evaluation coefficient T1 to determine the evaluation result.

[0029] Furthermore, the inter-annual correlation of the earth-rock dam safety evaluation index is obtained, and the inter-annual correlation is combined with the evaluation coefficient T1 to determine the evaluation result as follows:

[0030] Calculate the membership degree of the safety evaluation index in year a, and establish the original data matrix T based on the calculated membership degree;

[0031] Normalize the original data matrix and construct the fuzzy judgment matrix R;

[0032] According to the fuzzy judgment matrix, the data between different years are calculated using the quantity product method to obtain the similarity λ between different years;

[0033] The similarity λ is compared with a preset threshold, and if the similarity exceeds the threshold, it is determined that there is a turning point in that year;

[0034] Obtain the membership degree of each safety evaluation indicator in that year and determine whether it is dangerous based on the membership degree.

[0035] The beneficial effects of the present invention are:

[0036] (1) The present invention establishes an evaluation system based on the failure modes of earth-rock dams and their locations, which conforms to the actual situation;

[0037] (2) This invention addresses the problem of strong subjectivity in indicator designation and weight determination in existing comprehensive evaluation methods and adopts game theory to dynamically adjust and modify the weights;

[0038] (3) The present invention uses a clustering method to cluster the results between years and gives a safety evaluation result based on the comprehensive subjective and objective conclusions. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of the process of the present invention.

[0040] Figure 2 Schematic diagram of the safety evaluation index system for layered earth-rock dams in the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0042] like Figure 1 As shown in FIG, a clustering earth-rock dam evaluation method based on sub-items and sub-locations includes the following steps:

[0043] Step 1: Construct a safety evaluation index system for earth-rockfill dams by item and location;

[0044] According to existing specifications and failure modes of earth-rockfill dams, a comprehensive health diagnosis and evaluation system for the safety of high earth-rockfill dams is constructed based on different parts and items, including the dam crest, wave-breaking wall, upstream slope protection, water surface near the dam, downstream slope protection, drainage system, dam foundation, and dam shoulders on both sides, according to sub-items such as flood resistance, structural stability, seepage stability, seismic safety, and metal structure. This is the layered earth-rockfill dam safety indicator evaluation system of the present invention.

[0045] Safety index evaluation system for layered earth-rock dams Figure 2 Shown are:

[0046] The earth-rock dam safety evaluation set is denoted as C i ={c i1 , c i2 ,…,c im}, C i is the project layer, n is the number of projects; C i ={c i1 , c i2 ,…,c im}, where c ij is the jth evaluation index of the i-th layer, and m is the number of indicators;

[0047] The project layer includes flood resistance, structural stability, seepage stability, seismic safety, and metal structure. Flood resistance includes two indicators: dam crest height verification and downstream flood discharge verification. Structural stability includes the following indicators: dam slope stability, dam body deformation, impermeable body deformation, other structural deformation, contact deformation, and near-dam bank slope. Seepage stability includes the following indicators: seepage pressure, seepage gradient, seepage volume, and seepage water quality. Seismic safety includes the following indicators: dam seismic safety factor and soil liquefaction potential. Metal structure includes the operation of steel gates and hoists.

[0048] Step 2: Determine the weights of subjective and objective indicators for each item and each part; determine the objective indicator weight ω2 based on the monitoring data, and determine the subjective indicator weight ω1 based on the preset expert experience;

[0049] Subjective weights are obtained using existing preset empirical values. Generally, experts are invited to obtain all indicator weight libraries based on their own experience according to various situations, and empirical values ​​are obtained according to the situation. The specific determination process is as follows:

[0050] Establish an evaluation table for indicators and projects, issue expert scores, and obtain the weights of the project and indicator layers. ij (c i ), c ik (c j ) to compare them pairwise and obtain their relative importance value α ijk (α ij ), forming a weight judgment matrix A=(α ijk ) m×m Or A=(α ij ) n×n The obtained importance indices are shown in Table 1.

[0051] Table 1. Importance score comparison table

[0052]

[0053] According to the weight judgment matrix A, the maximum eigenvalue λ of the judgment matrix is ​​obtained max , according to the characteristic equation AW=λ max W is calculated to obtain the eigenvector W, and the indicator weight can be obtained after normalizing the eigenvector.

[0054] The calculated indicator weights are checked for consistency. The function for judging the consistency of the matrix is ​​as follows:

[0055]

[0056] When R=0, the matrix is ​​consistent. The larger the R value is, the higher the degree of inconsistency of the judgment matrix A is.

[0057] Using subsamples, the critical values ​​of R obtained for different n were selected, and the results are shown in Table 2.

[0058] Table 2. Critical value selection

[0059]

[0060] The objective indicator weight ω2 is determined according to the entropy weight method, and the process is as follows:

[0061] The average amount of information after excluding redundancy is called "information entropy". According to the information entropy theory, the index c ij The information entropy of is calculated as follows:

[0062]

[0063] Divide the information into n intervals and calculate the index c based on the K intervals ij The membership degree is denoted as C ijk . By information entropy E ij The determined evaluation index c ij The entropy weight calculation process is as follows:

[0064]

[0065] Step 3: Adjust the weights according to the game theory method to obtain the adjusted combined weight ω;

[0066] ω=δ k ω1+(1-δ k )ω2

[0067] Among them, δ k is the weight distribution coefficient of objective weight and subjective weight, δ k Determined based on game theory methods.

[0068] Comprehensive safety evaluation index system for high earth-rock dams For example, when the monitoring data of a certain part changes little and the trend is stable, it means that the structural safety status here is good. At this time, the ω2 obtained through the monitoring data is small and the change is stable, which means the attention is small. k The subjective importance of the structure itself is mainly assigned; when the monitoring value fluctuates or even rises significantly, ω2 is large and the change is stable, it means that the structure is unstable and needs to be paid more attention. At this time, it is necessary to give more attention to the objective weight. The traditional method mainly uses artificial assignment to balance the subjective and objective weights. This method introduces game theory to calculate δ k To adjust the ratio between subjective and objective.

[0069] Game theory is a powerful tool for solving competitive problems. It can be used to analyze the rational behavior of multiple decision-makers and the equilibrium of their decisions when their actions influence each other. The basic idea behind using game theory to adjust weights based on changes in subjective and objective sequences is to find the equilibrium point between the results of different weightings, minimizing the sum of the deviations between the combined weight and the results of various weightings, thereby selecting the optimal combined weight.

[0070] Assuming that there are a total of r methods for solving weights, the number of evaluation indicators is m, and the corresponding weight vector construction method, the constructed minimum deviation optimization model is shown in the following formula. The model integrates the weight information of various weighting methods and can achieve the unification of subjective and objective weighting methods.

[0071]

[0072] The above model is solved by the Lagrangian function, which is as follows:

[0073]

[0074] Among them: λ is the introduced parameter, respectively Taking the derivative of each component with λ, we can simplify it to get:

[0075]

[0076] The above formula contains an r+1-dimensional linear system of equations with unknown variables. If the coefficient determinant is not equal to zero, it has a unique solution. The solution vector is the weight distribution coefficient for each weighting method. The final combined weight value is obtained by adding the indicator weights obtained by various weighting methods. The basic weighting method used is to calculate the objective weights of each level of indicators based on monitoring data and determine the subjective weights of each level based on expert opinions. Therefore, r = 2.

[0077] Step 4: Determine the health grade of the earth-rockfill dam based on the experience of large dam evaluation and regular inspection results;

[0078] According to the "Regulations on the Safety Management of Dam Operation of Hydropower Stations", dam safety levels are divided into three levels: normal dam, sick dam and dangerous dam. As shown in Table 3:

[0079] Table 3. Dam safety levels according to the regulations on dam operation safety management of hydropower stations

[0080]

[0081]

[0082] The intervals specified in the present invention are:

[0083] N = {[1, 0.8], (0.8, 0.6], (0.6, 0]} = {normal dam (class 1 dam), sick dam (class 2 dam), dangerous dam (class 3 dam)}

[0084] Where N is the classification of earth-rockfill dams. For high earth-rockfill dams, key indicators are crucial for dam safety. However, the calculation model based on a system of indicators can easily overwhelm the impact of a single indicator. Therefore, key indicators for stability and seepage damage are identified. When an indicator is 0, a global adjustment coefficient δT of 0.6 is introduced. When a particular indicator is within the warning range, the adjustment coefficient δT is 0.8. For other adjustment coefficients, δT is 1, corresponding to the dam level. These adjustment coefficients are used to calculate the evaluation coefficient. Here, the indicators in step 1 are quantified before judgment is made. The quantification method follows existing methods.

[0085] Step 5: Calculate the membership degree of each indicator based on the monitoring data;

[0086] The membership degree is calculated based on the project level and specific safety evaluation indicators.

[0087] (1) Flood resistance

[0088] Flood control capacity is assessed by reviewing design floods and performing flood control calculations based on hydrological data during operation, to determine whether the dam's flood control capacity meets existing standards and flood control requirements. Key indicators include elevation review and downstream flood review.

[0089] The membership function of the elevation review index is:

[0090]

[0091] Where: h 实 is the actual dam crest elevation, h 规 It is the dam crest elevation required by current specifications.

[0092] Flood discharge capacity membership:

[0093]

[0094] Where: Q 泄 To calculate the maximum possible discharge of the dam, Q 允 The maximum discharge flow allowed downstream.

[0095] (2) Structural stability

[0096] The purpose of structural safety assessment is to review whether the deformation, strength and stability of the dam (including the reservoir bank near the dam) under static conditions meet the requirements according to the current national standards.

[0097] 1) Anti-slip and stable

[0098] The membership degree of anti-sliding stability is:

[0099]

[0100] Where: K is the dam slope stability safety factor calculated based on the measured pressure, K A , K C is the dam slope stability threshold, determined according to current specifications.

[0101] 2) Dam deformation

[0102] The membership degree of the deformed evaluation is:

[0103]

[0104] Where: s is the measured value of deformation, s a 、s c is the deformation threshold.

[0105] According to the regulations, the maximum settlement of the earth-rock dam crest after completion shall not exceed 1% of the dam height. This value can be used as the upper critical value, and the lower critical value can be included in a certain redundancy. The present invention initially proposes to multiply the upper limit allowable value by a reduction coefficient of 0.9, which can be appropriately adjusted for different projects.

[0106] There is no specific control standard for deformation during construction. Existing technology has derived an empirical relationship between the settlement S and the dam height H after the rockfill dam project reaches stability:

[0107]

[0108] Where: S is the settlement and H is the dam height.

[0109] Assuming that the deformation of the rockfill body due to its own weight has been completely eliminated during the construction period, the diffusion of water pressure in the dam is linear, and the relationship between stress and deformation is linear, the vertical deformation prediction formula of the rockfill dam is:

[0110] S=0.0001245H 2

[0111] The above-mentioned rockfill dam deformation prediction formula focuses on vertical deformation and does not include estimation methods for several other deformations. Therefore, this paper uses numerical simulation combined with engineering statistics to establish a formula for predicting the maximum deformation of the dam surface during the construction period of an earth-rockfill dam. The three-dimensional finite element numerical simulation scheme is shown in Table 4:

[0112] Table 4. Three-dimensional finite element numerical simulation of earth-rock dam deformation

[0113]

[0114]

[0115] The soil constitutive model adopts the Duncan-Zhang EB model, and the material parameters are shown in Table 5.

[0116] Table 5. Material parameters of three-dimensional finite element numerical simulation scheme for earth-rockfill dam deformation

[0117]

[0118]

[0119] Based on the numerical simulation results of multiple schemes and the deformation monitoring results of several high core rockfill dams, the following empirical formula for calculating the maximum deformation of the dam surface was obtained through statistical fitting of the maximum deformation of the dam surface:

[0120] Maximum vertical deformation of dam surface:

[0121] y=215.94+2.859H+1.706H d

[0122] The maximum downstream horizontal deformation of the dam surface is:

[0123] y=48.852+1.465H+0.362H d

[0124] Maximum horizontal deformation of the dam surface:

[0125] y=±(-7.934+1.08H+0.431H d )

[0126] 3) Deformation of the anti-seepage body

[0127] For the soil core wall, the membership function can be established by considering its arch effect. The calculation formula is:

[0128]

[0129] Where: G is the arch effect coefficient, σ is the measured soil pressure, γ is the soil bulk density, and h is the dam height.

[0130] The foundation anti-seepage wall is mostly made of concrete anti-seepage wall, and the membership function can be established according to the strength:

[0131]

[0132] Where: T is the actual strength measured by the steel bar meter or stress gauge, T c is the critical value of strength danger, which is initially planned to be the strength value multiplied by the reduction factor of 0.9, T a The allowable strength value of the material.

[0133] 4) Other structural deformation

[0134] Other structures of earth-rock dams mainly include observation corridors, water diversion and flood discharge facilities, etc. These structures are mostly made of concrete and are equipped with steel bar gauges and strain gauges. Their membership functions can be established based on strength standards, and the calculation method is such as C23.

[0135] 5) Contact deformation

[0136] Contact deformation is determined by analyzing whether there is abnormal deformation and whether the deformation trend converges. Its membership function is:

[0137]

[0138] Where: n 异 To increase the confidence level of abnormal deformation and trend judgment, abnormal deformation is identified using the structural anomaly recognition method, and deformation trends are identified using the first-order and second-order derivatives.

[0139] 6) Slope deformation near the dam

[0140] The safety of the bank slope near the dam area can be calculated by calculating the corresponding index membership through slope stability calculation and slope deformation rate. The slope stability membership is the same as the dam stability membership. The slope deformation rate membership function is:

[0141]

[0142] Where: c is the deformation rate monitored by the multi-point displacement meter, c a is the empirical allowable value of deformation rate, determined based on experience.

[0143] The construction period of rock slope is 1mm / d, and the operation period is 0.5mm / d; the construction period of soil slope is 5mm / d, and the operation period is 1mm / d; c c In order to consider the critical value of redundancy deformation, it is initially proposed to multiply the allowable value by a reduction factor of 0.9.

[0144] (3) Stable seepage

[0145] The purpose of seepage stability evaluation is to review whether the seepage control measures originally designed and constructed and the current actual seepage status can ensure the safe operation of the dam according to the design conditions.

[0146] 1) Seepage pressure

[0147] The permeability pressure can be compared with the design value to establish the membership function:

[0148]

[0149] Where: P 实 is the measured osmotic pressure value, P 设 is the design value for seepage.

[0150] 2) Infiltration gradient

[0151] The membership of the seepage slope is determined by comparing the calculated value of the seepage measurement point with the allowable seepage slope of the soil:

[0152]

[0153] Where: i is the infiltration slope, i a is the permissible infiltration slope; i c In order to consider the seepage slope threshold value of the redundancy, it is initially proposed to multiply the allowable value by a reduction factor of 0.95.

[0154] 3) Seepage

[0155] The seepage rate membership function can be established by comparing the monitored value with the design value:

[0156]

[0157] Where: Q is the seepage volume of the dam, Q a is the design seepage rate, Q c In order to consider the critical value of redundancy seepage, it is initially proposed to multiply the allowable value by a reduction factor of 0.9.

[0158] 4) Water quality

[0159] Water quality is mainly determined through manual inspections and experiments, and qualitative conclusions are drawn:

[0160]

[0161] (4) Seismic stability

[0162] The purpose of a seismic safety review of an earth-rockfill dam is to verify whether the current status of the dam project meets seismic requirements according to current standards. The review targets permanent retaining structures, spillway and water transfer structures related to dam safety, as well as foundations and the near-dam bank.

[0163] 1) Dam seismic safety

[0164] The dam slope stability can be obtained by using the pseudo-static method or finite element dynamic calculation considering earthquake acceleration, and the membership function is the same as that of the dam slope stability.

[0165] 2) Soil liquefaction safety

[0166] Soil liquefaction is determined based on soil type, SPF number, relative density, soil property index, dynamic triaxial test, dynamic shear strength, etc. The membership function can be determined based on SPF test:

[0167]

[0168] Where: N is the standard penetration point d below the ground when used in engineerings Number of penetration blows at depth, N cr The standard penetration hammer number for liquefaction judgment.

[0169] (5) Metal structure

[0170] The purpose of metal structure safety assessment is to verify whether the steel gates, hoists and pressure pipes of reservoir dam discharge and water delivery structures can operate safely according to design conditions under the current situation.

[0171] 1) Working condition of steel gate

[0172] The steel gate is mainly checked to see whether its strength and stiffness meet the requirements. Its membership function is calculated in the same way as the evaluation coefficient T1.

[0173] 2) Working condition of gate hoist

[0174] The working condition of the gate hoist mainly analyzes whether the gate hoist's opening and closing capacity meets the requirements. Its membership function is

[0175]

[0176] Where: q is the opening and closing capacity of the gate hoist, and L is the opening and closing load.

[0177] Another review of the gate hoist is to evaluate whether it can ensure the normal opening and closing of the equipment in an emergency. This is evaluated by engineers and experts.

[0178] Step 6: Obtain the evaluation coefficient based on the membership degree obtained in step 5, and combine it with the health level in step 4 to obtain the evaluation result.

[0179] The evaluation coefficient calculation process is as follows:

[0180] Calculation project layer C i Membership degree x i :

[0181]

[0182] Where: x ij is the indicator c ij The corresponding membership degree, ω ij is the indicator c ij The corresponding weight;

[0183] Evaluation coefficient T1 is

[0184]

[0185] Where: δT is the adjustment coefficient, which is determined based on the safety evaluation index of earth-rock dam.

[0186] In order to make the evaluation more accurate, the clustering method is used to obtain the inter-annual correlation of the earth-rock dam safety evaluation indicators. The inter-annual correlation is combined with the evaluation coefficient T1 to determine the evaluation results. The process is as follows:

[0187] Set up earth-rock dam safety assessment project layer c i Indicator c ij The membership degree in year a is X ija , abbreviated as x ja , establish the original data matrix T;

[0188]

[0189] The original data matrix is ​​normalized by translation and extreme value transformation to construct the fuzzy judgment matrix R; different years are compared and the similarity τ is calculated by the quantitative product method between different years. ij =F(x i ,x j ); F is the similarity discriminant function, where x i is a row vector in T, x j is the column vector in T, and the Euclidean distance method is used to calculate:

[0190] τ ij =1-βd(x i ,x j )

[0191] Among them, β is a parameter, d is the distance, and the calculation process is as follows:

[0192]

[0193] The fuzzy similarity matrix established above is not necessarily transitive. Generally, the transitive closed-form method is used to transform the matrix to generate the final fuzzy equivalent matrix, and then the clustering relationship over time is judged based on the distance λ. Where λ = τ ij .

[0194] The similarity λ is compared with a preset threshold, and if the similarity exceeds the threshold, it is determined that there is a turning point in that year;

[0195] Obtain the membership degree of each safety evaluation indicator in that year and determine whether it is dangerous based on the membership degree.

[0196] The present invention will be further described below with reference to specific embodiments.

[0197] According to a specific project, judge according to the above method.

[0198] Step 1: Construct a safety evaluation index system for earth-rockfill dams by item and location;

[0199] Step 2: Determine the weights of subjective and objective indicators for each item and each part; determine the objective indicator weight ω2 based on the monitoring data, and determine the subjective indicator weight ω1 based on the preset expert experience;

[0200] Step 3: Adjust the weights according to the game theory method to obtain the adjusted combined weight ω;

[0201] The subjective weights, objective weights, and comprehensive weights adjusted using game theory are shown in Table 6. Among the subjective weights, anti-sliding stability (C21), dam body deformation (C22), and anti-seepage deformation (C23) account for the highest proportions, at 0.22, 0.29, and 0.13, respectively. Other dam ancillary structure deformation (C24) accounts for a smaller proportion, at 0.15. Among the objective weights, since the calculated membership of factors such as dam slope stability has remained relatively unchanged over the past four years, the weight is relatively low, at 0.13. However, due to the significant changes in the membership of ancillary structures, the weight is relatively high, at 0.19. This calculation shows that information entropy can reflect the activity level of indicators. When certain indicators become less active, their weights decrease, while vice versa, they increase. This can serve as a beneficial balance to traditional weights.

[0202] Table 6. Subjective and objective weight table

[0203]

[0204] Step 4: Determine the health grade of the earth-rock dam based on the empirical coefficient of dam evaluation;

[0205] Step 5: Calculate the membership of each indicator based on the monitoring data; the present invention calculates the membership results from 2010 to 2013. The indicator determination process is shown in Table 7, and the calculated membership is shown in Table 8. For indicators such as seepage and deformation, the abnormality of a single measuring point cannot fully reflect the problem of the entire indicator. The following method is used to calculate multiple indicators:

[0206]

[0207] Where: ∑C ij C ij The membership degree is , and s is the instrument failure rate.

[0208] Table 7. Safety index membership calculation table

[0209]

[0210]

[0211] Table 8. Safety index membership calculation table

[0212]

[0213]

[0214] Step 6: Obtain the evaluation coefficient based on the membership degree obtained in step 5, and combine it with the health level in step 4 to obtain the evaluation result.

[0215] According to the comprehensive evaluation method, the scores from 2010 to 2014 were 0.975, 0.972, 0.981, and 0.984. The dam was in good operating condition and was a normal dam or a Class I dam. The scores continued to improve over time, resulting in a clustering result. The similarity of dam safety in 2010 and 2011 was 0.96; the similarity between 2012 and 2013 was 0.94, and the overall similarity was 0.81, indicating that the safety conditions in 2010 and 2011 were basically similar, and the safety conditions in 2012 and 2013 were basically similar. Analysis of the reasons: The dam's flood control indicators, seismic stability indicators, and metal structure indicators all met the requirements, with a membership degree of 1. Some measured values ​​of seepage pressure were high, but there were no major anomalies. Therefore, the clustering result was mainly affected by the structural stability indicator C. 22 、C 24 、C 25 In the observation corridor stress, cracks appeared in the corridor after 2008, and the stress exceeded the standard. Before 2011, the change rate was large, 0.02MPa / d. After 2011, the change value decreased to 0.01MPa / d, and has been basically stable so far. This has caused C 24 The indicators differed significantly before and after 2011. Dam deformation and contact deformation also showed significant changes during the construction and impoundment periods, but remained stable during operation. Therefore, 2011 can be used as a segmental boundary for the changes, which is consistent with the cluster analysis structure. The above evaluation indicates that the core wall dam is a normal or Class I dam, with overall stable operation and improved safety over time. After 2011, these indicators have become more normal.

[0216] Assuming the 2010 anti-sliding stability index is reduced to 0, it is clearly a dangerous dam or Class III dam according to the standard. The result calculated by the traditional method is 0.942, while the result obtained by the method of the present invention is 0.565, which is more in line with reality.

[0217] Compared with traditional evaluation methods, the method proposed in the present invention establishes an evaluation system based on the sub-items and parts of the earth-rock dam failure mode. The index system of the present invention is determined based on the historical experience of earth-rock dam failures, dam failure modes and relevant specifications. Common failure modes of earth-rock dams mainly include dam slope instability, core wall splitting, seepage damage to the dam body and foundation, and dam overflow. The paths of these dam failures are often manifested as the gradual expansion of local problems leading to overall failure. For example, local leakage may cause overall seepage damage in a certain part. At the same time, the patent adopts a sub-part and sub-item index system to more accurately capture the potential safety hazards of each part, and establishes a cluster diagnosis method to continuously and qualitatively evaluate the development process of hazards. By adopting this index system, through the push of point-line-surface-body, we can see the big picture from the small, identify risks early, and avoid evaluation blind spots.

[0218] This invention addresses the highly subjective nature of indicator formulation and weighting in comprehensive evaluation methods by employing game theory to dynamically adjust and modify weights, assigning objective values ​​based on interannual variations. For example, large, continuous objective changes indicate active objective information, and game theory therefore increases the importance of objective factors, consistent with common understanding. Furthermore, to address the issue of buried membership information, key indicators are globally adjusted. Finally, clustering is employed to combine subjective and objective conclusions to produce a safety diagnosis.

Claims

1. A clustered earth-rock dam safety assessment method based on item and location, characterized in that: The following steps are involved: Step 1: Construct a safety evaluation index system for earth-rock dams by item and location. The safety evaluation index system for earth-rock dams is as follows: The safety evaluation set of earth-rock dam is denoted as C = {C1, C2, ..., C n }, C i is the project layer, n is the number of projects; C i ={c i1 , c i2 ,…,c im }, where c ij is the jth evaluation index of the i-th layer, and m is the number of indicators; The project layer includes flood resistance, structural stability, seepage stability, seismic safety, and metal structure; flood resistance includes dam crest height review and downstream flood review; structural stability includes dam slope stability, dam body deformation, anti-seepage body deformation, other structural deformation, contact deformation, and bank slope near the dam; seepage stability includes seepage pressure, seepage gradient, seepage volume, and seepage water quality; seismic safety includes dam seismic safety factor and soil liquefaction possibility; metal structure includes steel gate operation and hoist operation. Step 2: Determine the weights of subjective and objective indicators for each item and each part; determine the objective indicator weight ω2 based on the monitoring data, and determine the subjective indicator weight ω1 based on the preset expert experience; Step 3: Adjust the weights according to the game theory method to obtain the adjusted combined weight ω; Step 4: Determine the health level of the earth-rockfill dam; Step 5: Calculate the membership degree of each indicator based on the monitoring data; Step 6: Obtain the evaluation coefficient based on the membership degree obtained in step 5, and use cluster analysis to obtain the evaluation results in combination with the health level in step 4; The evaluation coefficient calculation process is as follows: Calculation project layer C i Membership degree x i ; Where: x ij is the indicator c ij The corresponding membership degree, ω ij is the indicator c ij The corresponding weight; Evaluation coefficient T1 is Where: T is the adjustment coefficient, which is determined based on the safety evaluation index of earth-rock dam; The inter-annual correlation of earth-rock dam safety evaluation indicators is obtained and the evaluation results are determined by combining the inter-annual correlation with the evaluation coefficient T1 as follows: Calculate the membership degree of the safety evaluation index in year a, and establish the original data matrix T based on the calculated membership degree; Normalize the original data matrix and construct the fuzzy judgment matrix R; According to the fuzzy judgment matrix, the data between different years are calculated using the quantity product method to obtain the similarity λ between different years; The similarity λ is compared with a preset threshold, and if the similarity exceeds the threshold, it is determined that there is a turning point in that year; Obtain the membership degree of each safety evaluation indicator in that year and determine whether it is dangerous based on the membership degree.

2. The method for clustering earth-rock dam safety assessment based on item and location according to claim 1 is characterized in that: The calculation method of the combination weight is as follows: ω=d k ω1+(1-δ k )ω2 Among them, δ k is the weight distribution coefficient of objective weight and subjective weight, δ k Determined according to the game theory method, ω1 is the subjective indicator weight and ω2 is the objective indicator weight.

3. The method for clustering earth-rock dam safety assessment based on item and location according to claim 2 is characterized in that: The subjective indicator weight ω1 and the objective indicator weight ω2 are determined according to the entropy weight method.

Images