Method for predicting residual life of small reservoir based on health degree evaluation model

Abstract

The invention relates to the technical field of hydraulic engineering safety monitoring, in particular to a small reservoir residual life prediction method based on a health degree evaluation model, comprising the following steps: constructing a small reservoir health evaluation index system; calculating the health degree of the small reservoir sub-project; calculating the comprehensive health degree of the small reservoir; establishing a nonlinear Wiener process model; estimating model parameter values; predicting the residual life of the small reservoir; the beneficial effects are that according to the small reservoir residual life prediction method based on the health degree evaluation model provided by the invention, the comprehensive health degree value representing the health state of the small reservoir is calculated according to the health degree calculation result of each sub-project of the small reservoir; a degradation model based on a small reservoir health degree index is constructed by using a nonlinear Wiener process, a small reservoir residual life probability density function is deduced, and a small reservoir residual life value can be obtained more accurately.

Description

Technical Field

[0001] The invention relates to the field of water conservancy project safety monitoring, and in particular to a method for predicting the remaining life of a small reservoir based on a health assessment model. Background Art

[0002] In order to timely prevent and discover possible safety hazards in small reservoirs such as dam cracks, landslides, leakage, etc., and take preventive measures to avoid or reduce the occurrence of disasters, it is particularly important to carry out long-term monitoring of small reservoirs and grasp the changes in the safety status of small reservoirs.

[0003] In the prior art, the empirical method is often used in the prediction of the remaining life of small reservoirs, that is, the prediction is made based on the life and operation of similar reservoirs combined with the actual situation of the current reservoir. Or the life of the reservoir is predicted by establishing a mathematical model, considering multiple influencing factors, and performing complex calculations and analyses.

[0004] The above method mainly conducts health assessment by dividing the water into safety grades. It does not integrate the data of each grade to obtain a health degree that can comprehensively characterize the health status of small reservoirs. It cannot reflect the degradation status of small reservoirs in more detail and cannot provide an accurate basis for the situation-based maintenance of small reservoirs. Summary of the invention

[0005] The purpose of the present invention is to provide a method for predicting the remaining life of a small reservoir based on a health assessment model. The health status can more carefully reflect the degradation state of the small reservoir, provide an accurate basis for the situation-based maintenance of the small reservoir, and solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a method for predicting the remaining life of a small reservoir based on a health assessment model, the method comprising the following steps:

[0007] S1. Construct a small reservoir health assessment indicator system;

[0008] S2. Calculation of health of small reservoir subprojects;

[0009] S3. Calculation of comprehensive health of small reservoirs;

[0010] S4. Establish a nonlinear Wiener process model;

[0011] S5. Estimation of model parameter values;

[0012] S6. Prediction of remaining life of small reservoirs.

[0013] Preferably, the specific process of step S1 includes: according to the characteristics of safety monitoring of major structures of small reservoirs, following the principles of purpose, scientificity, systematicity and measurability in selecting health assessment indicators, a small reservoir health assessment indicator system is constructed.

[0014] Preferably, the specific process of step S2 includes:

[0015] S21. Division of sub-project indicator evaluation sets;

[0016] S22. Determination of the sub-project indicator affiliation;

[0017] S23. Determination of sub-project indicator weights;

[0018] S24. Calculation of sub-project health.

[0019] Preferably, the specific process of step S21 includes: combining existing engineering experience methods and the actual situation of the sub-project engineering, reasonably grading the evaluation indicators of the sub-projects, and directly determining the evaluation set by different characteristic values ​​of the indicators when there are clear grading characteristic values. For indicators without clear grading characteristic values, the evaluation set is divided by the confidence level method;

[0020] The specific process of step S22 includes: using fuzzy mathematics theory to determine the membership of each factor in the factor set to the evaluation level, summarizing the matrix obtained as the evaluation matrix, and using fuzzy statistics to determine the membership of indicators with clear classification characteristic values; for indicators without clear classification characteristic values, using normal distribution function as the membership function;

[0021] The specific process of step S23 includes: for the first-level indicators, the hierarchical analysis method is used to construct the weights, the "1-9 scaling method" is used to compare the elements in pairs to form a judgment matrix, and then the square root method is used to solve the weights, and finally the judgment matrix is ​​tested for consistency; for the second-level indicators, the entropy weights are calculated according to the degree of variation of each indicator, and then the weights are corrected by the entropy weights to obtain the objective weights of the indicators; through step-by-step evaluation, the fuzzy synthesis operation of the membership vectors of each indicator is performed to obtain the evaluation matrix.

[0022] Step S24 includes: proposing a comprehensive evaluation index F, with health assessment level V = {v1, v2, ... v n} as variables, and comprehensively judge the membership vector B = {b1, b2, ... b n} as the weight, and get the comprehensive evaluation index:

[0023]

[0024] Where: v j is the number of the corresponding level j; n is the total number of evaluation levels; k is the coefficient, which is 1 in the project;

[0025] Propose the health degree H and calculate the sub-project health degree as:

[0026]

[0027] Where: The health degree H can quantitatively describe the operating status of the sub-project from healthy H=1 to damaged H=0.

[0028] Preferably, the specific process of step S3 includes: using the health of each sub-project calculated in step S2 as an evaluation index for calculating the comprehensive health of the small reservoir, and calculating the comprehensive health of the small reservoir with reference to step S2.

[0029] Preferably, the specific process of step S4 includes: let X(t) represent the degradation amount of the small reservoir at time t, and establish a health degradation model of the nonlinear Wiener process:

[0030] X(t)=x(0)+a∫0 t μ(b,t)dt+σ B B(t)

[0031] Where: t is time; x(0) represents the degradation of a small reservoir at time 0. In engineering applications, x(0) = 0. a is the drift coefficient, σ B is the diffusion coefficient, B(t) is the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis; μ(b,t) and B(t) are independent of each other. is a nonlinear function with b as parameter, Different functional forms of can describe different forms of nonlinear random processes.

[0032] Preferably, the specific process of step S5 includes: establishing a nonlinear degradation model containing parameters θ(a, b, σ B ), for the deterministic trend term θ(a,b) containing unknown parameters, the nonlinear least squares method is used to estimate the unknown parameters, and the maximum likelihood estimation method is used to estimate the parameter σ B Make an estimate.

[0033] Preferably, the specific process of step S6 includes: according to the law that the first arrival time in the Wiener process obeys the inverse Gaussian distribution, the remaining life probability density function of the small reservoir health degradation model is obtained, and the remaining service life process of the small reservoir is obtained.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention proposes a method for predicting the remaining life of a small reservoir based on a health assessment model. The method calculates a comprehensive health value that characterizes the health status of the small reservoir according to the results of each level of the comprehensive membership of indicators of each sub-item in the evaluation of the small reservoir. The nonlinear Wiener process is used to construct a degradation model based on the health index of the small reservoir, and the probability density function of the remaining life of the small reservoir is derived, so as to more accurately obtain the remaining life value of the small reservoir. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flow chart of the method of the present invention;

[0037] Figure 2 This is a schematic diagram of the remaining life prediction principle of the present invention;

[0038] Figure 3 It is a schematic diagram of the comprehensive health assessment index system of a small reservoir of the present invention;

[0039] Figure 4 It is a curve diagram of reservoir bank slope health degradation of the present invention;

[0040] Figure 5 It is a schematic diagram of the probability density function of the remaining life of the reservoir slope of the present invention;

[0041] Figure 6 It is a schematic diagram of the probability density of the remaining life of the reservoir slope in each period of the model of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose and technical solution of the present invention clearly and completely described, and the advantages more clearly understood, the embodiments of the present invention are further described in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, rather than all of the embodiments, and are only used to explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] For example, see Figure 1 to Figure 2 The present invention provides a technical solution: a method for predicting the remaining life of a small reservoir based on a health assessment model, the method comprising the following steps:

[0044] 1. Construct a small reservoir health assessment indicator system

[0045] According to the characteristics of safety monitoring of major structures of small reservoirs, a health assessment indicator system for small reservoirs is constructed in accordance with the principles of purpose, scientificity, systematicness and measurability in the selection of health assessment indicators.

[0046] 2. Calculation of health of small reservoir subprojects

[0047] (1) Division of sub-project indicator evaluation sets

[0048] Combined with existing engineering experience and methods, and in connection with the actual situation of sub-project engineering, each evaluation index of the sub-project is reasonably graded. When there are clear graded characteristic values, the evaluation set can be directly determined by the different characteristic values ​​of the indicators. For indicators without clear graded characteristic values, the evaluation set is divided by confidence level.

[0049] (2) Determination of sub-project indicator affiliation

[0050] The fuzzy mathematical theory is used to determine the membership of each factor in the factor set to the evaluation level, and the matrix obtained by summarizing them is the evaluation matrix. For indicators with clear classification eigenvalues, the fuzzy statistical method is used to determine the membership, and the calculation formula is:

[0051]

[0052] In the formula: j is the evaluation level, u j is the evaluation factor belonging to the evaluation level v j The number of measurements; u is the total number of samples.

[0053] For indicators without clear classification characteristic values, the normal distribution function is selected as the membership function, and its expression is:

[0054]

[0055] In the formula: x is the monitoring value of the corresponding evaluation index; σ is the mean square error; a j is the median value of the jth level in the region.

[0056] (3) Determination of sub-project indicator weights

[0057] For the first-level indicators, the hierarchical analysis method is used to construct the weights and the hierarchical structure model of the evaluation indicators. The "1-9 scaling method" is used to compare the elements pairwise to form a judgment matrix. The weights are then solved using the square root method. Finally, the judgment matrix is ​​tested for consistency.

[0058] For the secondary indicators, according to the degree of variation of each indicator, the entropy weight is calculated using information entropy, and then the weight is corrected by the entropy weight to obtain the objective weight of the indicator. The basic calculation formula is as follows:

[0059]

[0060] Where: l = 1 / lnm; S j is the index entropy value; p ij is the specific data of the indicator; j is the indicator weight.

[0061] Through step-by-step evaluation, the fuzzy synthesis operation of the membership vectors of each indicator is performed to obtain the evaluation matrix B:

[0062] B=WR=(b1,b1,…,b n ) (5)

[0063] Where: W is the weight of each first-level indicator; R is the membership result of each first-level indicator.

[0064] (4) Calculation of sub-project health

[0065] According to the characteristics of the evaluation analysis method, the functional mapping relationship between the sub-project health H and the membership vector is:

[0066] H=f(x1,x2,…,x n ) (6)

[0067] Where: f is the sub-project health quantification function, x i It is the numerical value of the membership degree vector in the sub-project health assessment.

[0068] A comprehensive evaluation index F is proposed, with health assessment level V = {v1, v2, …v n} as variables, and comprehensively judge the membership vector B = {b1, b2, ... b n} as the weight, and get the comprehensive evaluation index:

[0069]

[0070] Where: v j is the number corresponding to the jth grade level; n is the total number of evaluation levels; k is the coefficient, which is generally taken as 1 in engineering.

[0071] The health degree H is proposed, and the formula for calculating the health degree of the sub-project is:

[0072]

[0073] Where: The health degree H can quantitatively describe the operating status of the sub-project from healthy H=1 to damaged H=0.

[0074] 3. Calculation of comprehensive health of small reservoirs

[0075] The health of each sub-project calculated in step 2 is used as the evaluation index for calculating the comprehensive health of the small reservoir, and the comprehensive health of the small reservoir is calculated with reference to step 2.

[0076] 4. Establish a nonlinear Wiener process model

[0077] Let X(t) represent the degradation amount of the small reservoir at time t. Considering that the degradation amount of the small reservoir usually has nonlinear characteristics as the external environment changes during the degradation process, the degradation model of the nonlinear Wiener process can be expressed as:

[0078]

[0079] Where: t is time; x(0) represents the degradation of a small reservoir at time 0. In engineering applications, x(0) = 0. a is the drift coefficient, σ B is the diffusion coefficient, and B(t) is the standard Brownian motion, which is used to describe the uncertainty of the random degradation quantity on the time axis. μ(b; t) and B(t) are independent of each other; It is a nonlinear function with b as parameter, which is used to characterize the nonlinear characteristics of health degradation of small reservoirs. Different functional forms of can describe different forms of nonlinear random processes.

[0080] 5. Model parameter value estimation

[0081] The established nonlinear degradation model contains parameters θ(a, b, σ B ), for the deterministic trend term θ(a,b) containing unknown parameters, the nonlinear least squares method is used to estimate the unknown parameters so that A in the following formula is minimized.

[0082]

[0083] make Z(t) can be regarded as a special case of a linear Wiener process with zero drift. The maximum likelihood estimation method is used to estimate the parameter σ B Make an estimate.

[0084] 6. Prediction of remaining life of small reservoirs

[0085] like Figure 2 , the small reservoir experiences continuous degradation during operation. When the degradation of health reaches the preset failure threshold w for the first time, it can be considered that the small reservoir is unstable. Based on the concept of first arrival time, the small reservoir is unstable at t i The remaining life at time t f Defined as:

[0086] t f = inf{t f :X(t f +t i )≥w|X(t f ) <w}(11)

[0087] Where: X(t i ) is the degradation amount of the health of the small reservoir at the current moment.

[0088] According to the law that the first arrival time in the Wiener process obeys the inverse Gaussian distribution, the remaining life probability density function of the small reservoir health degradation model is obtained, and the remaining service life process of the small reservoir is obtained.

[0089] Example 2: In view of the fact that small reservoirs generally have many sub-projects, in order to briefly introduce the prediction method of their remaining life, based on Example 1, it is assumed that it only contains one sub-project, the reservoir slope, and its remaining life is predicted. The specific steps are as follows:

[0090] 1. Construct a small reservoir health assessment indicator system

[0091] Determine the displacement, water level and rainfall that are closely related to the bank slope status as its evaluation indicators. There are a total of 6 automatic measurement points for bank slope displacement, and analyze the measured horizontal displacement data. For the water level index, it is obtained through the measured water level data of the project area, and for the rainfall index, it is obtained based on the measured rainfall data of the area where the bank slope is located. Determine the bank slope health assessment index system as follows Figure 3 As shown. A health assessment is now conducted for the period from December 1, 2019 to November 1, 2020, and the slope health assessment level is divided into four levels, namely safe, basically safe, relatively unsafe, and unsafe. December 1, 2019 is used as an example to illustrate the calculation process, and the same can be obtained for other evaluation periods.

[0092] 2. Calculation of health of small reservoir subprojects

[0093] The displacement index is graded according to the confidence interval estimation method. The displacement index k is set to k1=1, k2=2. The displacement membership R1 of the six displacement measurement points is obtained by the membership calculation method:

[0094]

[0095] The entropy weight method is used to determine the weight of each measuring point, and then the first-level index membership matrix of the slope displacement is obtained:

[0096] R′1=[0.125 0.349 0.363 0.234]

[0097] The water level index is divided according to the evaluation level and the probability distribution characteristics. The values ​​of k are k1=1, k2=1.5, and k3=2. The rainfall evaluation level is determined according to the rainfall characteristics of the project area. The membership matrix of water level R2 and rainfall R3 is obtained as follows:

[0098] R2=[0.316 0.264 0.102 0.316]

[0099] R3=[0.835 0.141 0.024 0]

[0100] In summary, the first-level index membership matrix R of the slope can be obtained:

[0101]

[0102] According to the characteristics of each indicator in engineering practice, the first-level indicator weight vector W is determined by the hierarchical analysis method as follows:

[0103] W = [0.572 0.158 0.270]

[0104] After fuzzy synthesis operation, the comprehensive evaluation of the bank slope is obtained as follows:

[0105] B=WR=(0.226,0.228,0.308,0.184)

[0106] According to the maximum membership principle, the final fuzzy comprehensive evaluation result B of the bank slope is the third level.

[0107] 3. Calculation of comprehensive health of small reservoirs

[0108] Since it is assumed that the small reservoir contains only one subproject, the health of the subproject is mainly calculated as the health of the reservoir. According to the calculation results of the comprehensive membership of the bank slope, the health of the bank slope on December 1, 2019 is obtained by calculating formula (7) and formula (8) H = 0.5478. Similarly, the bank slope health is calculated from December 2, 2019 to November 1, 2020, and the calculation is performed every 5 days. The degradation curve of the bank slope health is as follows: Figure 4 .

[0109] 4. Establish a nonlinear Wiener process model

[0110] This embodiment selects to apply the power function form to the small reservoir project, and obtains the degradation model expression as follows:

[0111] X(t)=at b +σ B B(t)

[0112] 5. Model parameter estimation

[0113] The parameter estimation method of Example 1 is used to estimate the parameters of the model. The results are shown in Table 1.

[0114] Table 1 Parameter estimates of nonlinear Wiener model

[0115] parameter a b <![CDATA[σ B ]]> Estimated value 0.0019 1.044 <![CDATA[2.55*10 -4 ]]>

[0116] 6. Prediction of remaining life of small reservoirs

[0117] The remaining life probability density function of the reservoir degradation model is obtained:

[0118]

[0119] in: The probability density function of the remaining life of the reservoir is obtained, and its curve is as follows: Figure 5 .

[0120] To further verify the stability of the model, October 12, October 17, October 22, and October 22, 2020 were taken as the starting points for the remaining life prediction of the reservoir, and the probability density functions of the remaining life corresponding to each period were calculated using the health degradation curve as follows: Figure 6 shown.

[0121] Similarly, the probability density function is used to obtain the corresponding remaining life prediction results for each time period and they are listed in Table 2.

[0122] Table 2 Remaining life prediction results for each period

[0123] <![CDATA[Current time t i > <![CDATA[Remaining life prediction value t f > Predicting the time when a reservoir may become unstable October 12, 2020 803 December 24, 2022 October 17, 2020 793 December 19, 2022 October 22, 2020 799 December 30, 2022 October 27, 2020 787 December 23, 2022 November 1, 2020 785 December 26, 2022

[0124] From the calculation and analysis in Table 2, it can be seen that the remaining life prediction results of the model in different periods are very close, the model prediction results are stable, and in theory it can effectively predict the remaining life of small reservoirs.

[0125] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the remaining life of a small reservoir based on a health assessment model, characterized in that: The method comprises the following steps: S1. Construct a small reservoir health assessment indicator system; S2. Calculation of health of small reservoir subprojects; S3. Calculation of comprehensive health of small reservoirs; S4. Establish a nonlinear Wiener process model; S5. Estimation of model parameter values; S6. Prediction of remaining life of small reservoirs.

2. According to claim 1, a method for predicting the remaining life of a small reservoir based on a health assessment model is characterized by: The specific process of step S1 includes: according to the characteristics of safety monitoring of the main structures of small reservoirs, following the principles of purpose, scientificity, systematicness and measurability in selecting health assessment indicators, a small reservoir health assessment indicator system is constructed.

3. According to claim 1, a method for predicting the remaining life of a small reservoir based on a health assessment model is characterized by: The specific process of step S2 includes: S21. Division of sub-project indicator evaluation sets; S22. Determination of the sub-project indicator affiliation; S23. Determination of sub-project indicator weights; S24. Calculation of sub-project health.

4. According to the method for predicting the remaining life of a small reservoir based on a health assessment model according to claim 3, the specific process of step S21 includes: Combined with existing engineering experience and methods, and in connection with the actual situation of sub-projects, reasonable grading of sub-project evaluation indicators is carried out. When there are clear grading characteristic values, the evaluation set can be directly determined by the different characteristic values ​​of the indicators. For indicators without clear grading characteristic values, the evaluation set is divided by the confidence level. The specific process of step S22 includes: using fuzzy mathematics theory to determine the membership of each factor in the factor set to the evaluation level, summarizing the matrix obtained as the evaluation matrix, and using fuzzy statistics to determine the membership of indicators with clear classification characteristic values; for indicators without clear classification characteristic values, using normal distribution function as the membership function; The specific process of step S23 includes: for the first-level indicators, the hierarchical analysis method is used to construct the weights, the "1-9 scaling method" is used to compare the elements in pairs to form a judgment matrix, and then the square root method is used to solve the weights, and finally the judgment matrix is ​​tested for consistency; for the second-level indicators, the entropy weights are calculated according to the degree of variation of each indicator, and then the weights are corrected by the entropy weights to obtain the objective weights of the indicators; through step-by-step evaluation, the fuzzy synthesis operation of the membership vectors of each indicator is performed to obtain the evaluation matrix. Step S24 includes: proposing a comprehensive evaluation index F, with health assessment level V = {v1, v2, ... v n } as variables, and comprehensively judge the membership vector B = {b1, b2, ... b n } as the weight, and get the comprehensive evaluation index: Where: v j is the number of the corresponding level j; n is the total number of evaluation levels; k is the coefficient, which is 1 in the project; Propose the health degree H and calculate the sub-project health degree as: Where: The health degree H can quantitatively describe the operating status of the sub-project from healthy H=1 to damaged H=0.

5. According to the method for predicting the remaining life of a small reservoir based on a health assessment model according to claim 1, the specific process of step S3 comprises: The health of each sub-project calculated in step S2 is used as an evaluation index for calculating the comprehensive health of the small reservoir, and the comprehensive health of the small reservoir is calculated with reference to step S2.

6. According to the method for predicting the remaining life of a small reservoir based on a health assessment model according to claim 1, the specific process of step S4 includes: Let X(t) represent the degradation of the small reservoir at time t, and establish the health degradation model of nonlinear Wiener process: X(t)=x(0)+a∫0 t μ(b,t)dt+σ B B(t) Where: t is time; x(0) represents the degradation of a small reservoir at time 0. In engineering applications, x(0) = 0. a is the drift coefficient, σ B is the diffusion coefficient, B(t) is the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis; μ(b,t) and B(t) are independent of each other. is a nonlinear function with b as parameter, Different functional forms of can describe different forms of nonlinear random processes.

7. According to the method for predicting the remaining life of a small reservoir based on a health assessment model according to claim 1, the specific process of step S5 comprises: The established nonlinear degradation model contains parameters θ(a, b, σ B ), for the deterministic trend term θ(a,b) containing unknown parameters, the nonlinear least squares method is used to estimate the unknown parameters, and the maximum likelihood estimation method is used to estimate the parameter σ B Make an estimate.

Images