CN116167240A - Multi-measuring-point combined monitoring method for dam structure damage - Google Patents
Multi-measuring-point combined monitoring method for dam structure damage Download PDFInfo
- Publication number
- CN116167240A CN116167240A CN202310211580.8A CN202310211580A CN116167240A CN 116167240 A CN116167240 A CN 116167240A CN 202310211580 A CN202310211580 A CN 202310211580A CN 116167240 A CN116167240 A CN 116167240A
- Authority
- CN
- China
- Prior art keywords
- residual
- function
- monitoring
- dam
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Computation (AREA)
- Algebra (AREA)
- Geometry (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses a multi-measuring point combined monitoring method for dam structure damage, which comprises the following steps: firstly, a prediction model is independently built for each measuring point, residual error values of an actual measurement value and the prediction value are obtained, edge distribution of a residual error sequence is fitted, a Vine Copula model for calculating multi-dimensional joint probability distribution is determined, then the obtained corresponding edge distribution of new monitoring data residual error is substituted into the model to obtain a joint accumulated distribution function value at each moment, and whether damage exists in a structure is judged through comparison with an abnormal threshold value; according to the invention, through analyzing the correlation among residual sequences of a plurality of measuring points, a VineCopula model is adopted to fit the joint probability distribution of the residual sequences, and the obtained distribution cumulative function value is compared with a threshold value, so that the damage possibly occurring in a dam structure is monitored, and the sensitivity of static monitoring to the identification of structural local damage is improved.
Description
Technical Field
The invention relates to the technical field of dam body structural damage monitoring, in particular to a multi-measuring-point combined monitoring method for dam structural damage.
Background
The reservoir dam has multiple functions of flood control, power generation, irrigation, water supply and the like, is an important infrastructure of national economy, and under the influence of factors such as environmental change, load effect, natural disasters and the like, the phenomenon that the structural part or whole of the dam project is damaged and the bearing capacity tends to be deteriorated in the running process possibly occurs, so that the safety, applicability and durability of the structure are continuously reduced, the safety state of the dam structure which is often influenced by an inconspicuous slow continuous deterioration mechanism is monitored, the abnormal and damaged conditions in the structure are recognized as early as possible, and the potential safety hazard is found timely, so that the dam has important significance for guaranteeing long-term stable and safe running.
At present, most reservoir dams often adopt static structure health monitoring to achieve the purpose of detecting structural damage or degradation, because the engineering operation period is influenced by numerous environmental changes and load actions, the response behaviors of single sensor monitoring amounts to different factors can be known by constructing a data-based prediction model, including a statistical model adopting multiple linear regression or stepwise regression, an artificial neural network based on machine learning, an extreme learning machine, a support vector machine model and the like, on the basis, the traditional static monitoring is used for determining monitoring indexes by adopting a confidence interval method aiming at single measuring points, the prediction interval of future measured values is defined as a warning limit value, and new data of each sensor is independently evaluated according to the prediction interval, so as to judge whether to send an alarm or not.
However, although the abnormal condition of any sensor response variable monitoring value in the dam structure can be obtained by the single-measuring-point method, whether the damage exists in the whole dam structure or not can be reasonably judged after one-by-one analysis, the mutual connection between different measuring points in the whole dam structure cannot be reflected, the overall health state of the structure cannot be truly reflected, particularly, when the damage range of the structure is small and the degree is light, obvious change of sensor monitoring data cannot be caused generally, in this case, the single-measuring-point monitoring-based method cannot effectively monitor the abnormality of the structure, and therefore, the invention provides a multi-measuring-point combined monitoring method for the damage of the dam structure to solve the problems in the prior art.
Disclosure of Invention
Aiming at the problems, the invention aims to provide a multi-measuring point combined monitoring method for dam structural damage, which solves the problems that the existing static structural health monitoring method for the reservoir dam cannot reflect the interrelation between different measuring points in the whole dam, and cannot truly reflect the whole health state of the structure, so that the structural abnormality of the reservoir dam cannot be effectively monitored.
In order to achieve the purpose of the invention, the invention is realized by the following technical scheme: a multi-measuring point joint monitoring method for dam structure damage comprises the following steps:
step one: firstly, arranging monitoring points at different positions of a dam to be monitored to monitor the appearance deformation of the dam, then respectively fitting a corresponding prediction model for each measuring point according to monitoring data obtained by measuring the monitoring points arranged in the dam, and then calculating residual errors between actual measurement values of each monitoring point and prediction values of each prediction model to obtain each residual error sequence;
step two: firstly, selecting Kendall rank correlation coefficients as a measurement index for measuring residual sequence dependence to analyze the correlation and consistency among residual sequences, then selecting a Vine structure with the largest sum of Kendall rank correlation coefficient absolute values by using a maximum tree generation algorithm, and fitting a probability density function of residual sequence distribution by using a kernel density estimation method to obtain edge distribution of each residual sequence;
step three: firstly, distributing and selecting a binary Copula function for each pair of edges in a Vine structure in the second step according to a red pool information criterion, estimating parameters of the binary Copula function selected for each edge in the Vine by using a maximum likelihood estimation method, and constructing a Vine Copula model according to the selected Copula function and the selected binary Copula function;
step four: firstly, calculating the difference between the currently acquired monitoring data and the predicted value of the predicted model, obtaining a current residual sequence, substituting the current residual sequence into the probability density function in the second step to obtain edge distribution, and substituting the edge distribution into the Vine Copula model in the third step after judging to obtain a corresponding combined cumulative distribution function value at each moment;
step five: firstly determining an abnormal threshold value, then comparing the combined cumulative distribution function value obtained in the fourth step with the abnormal threshold value, and judging whether damage exists in the dam structure according to a comparison result.
The further improvement is that: in the second step, when the correlation and consistency between residual sequences are analyzed, X { X }, is set 1 ，x 2 ，L，x n Sum Y { Y } 1 ，y 2 ，L，y n The residual sequence of two measuring points (x) i ，y i ) And (x) j ，y j ) Residual data pairs respectively of points at two sides of i and j moments, if x i ＜x j And y is i ＜y j Or x i ＞x j And y is i ＞y j Then call (x) i ，y i ) And (x) j ，y j ) If x is consistent with i ＜x j And y is i ＞y j Or x i ＞x j And y is i ＜y j Then call (x) i ，y i ) And (x) j ，y j ) Non-uniform, when X, Y two sides are shared
A different residual data pair, kendall rank correlation coefficient, is defined as:
where c represents a consistent residual function pair and d represents a non-consistent residual data pair.
The further improvement is that: in the second step, the formula of the maximum tree generation algorithm is as follows:
wherein T is i T is the set of all possible tree structures in the ith tree i Representing the structure of a particular ith tree, e is tree t i Any edge, delta i.j And Kendall rank correlation coefficients of a pair of residual sequences corresponding to the e-side are represented.
The further improvement is that: in the second step, the probability density function has the expression:
in the method, in the process of the invention,
to fit the probability density function, x i K is a Gaussian kernel function, h is a bandwidth and is selected according to the optimization of the integral mean square error.
The further improvement is that: in the third step, the formula for calculating the erythro pool information criterion is as follows:
AIC＝2K-2ln(L)
wherein K is the number of parameters, for five Copula functions of Frank Copula, clayton Copula, gumbel Copula, gaussian Copula and t-Copula, K is 2 except for the t-Copula function, the other types are 1, and L is the maximum value of the likelihood function.
The further improvement is that: in the third step, the calculation formula of the likelihood function is as follows:
where n represents the number of samples, θ= (θ) 1 ，θ 2 ，…θ k ) The parameter vector representing the Vine Copula model, C (·) is the density function of the Copula function C (·), and F (·) is the edge distribution.
The further improvement is that: in the fourth step, the edge cumulative distribution function value judging formula is:
where CDF is the value of the edge cumulative distribution function.
The further improvement is that: in the fifth step, when the combined cumulative distribution function value is compared with the abnormal threshold value, if the combined cumulative distribution function value is lower than the abnormal threshold value, the dam structure is abnormal, and if the combined cumulative distribution function value is higher than the abnormal threshold value, the dam structure is normal.
The beneficial effects of the invention are as follows: after the prediction model of each measuring point is constructed, the monitoring interval of a single measuring point is not directly determined by adopting a confidence interval method, the correlation among residual sequences of a plurality of measuring points is considered, the joint probability distribution of the measuring points is fitted based on a Vine Copula model, the distribution cumulative function (CDF) value is used as an index for judging abnormality, and the CDF value of the joint probability distribution obtained by calculation is compared to monitor the damage possibly occurring in a dam structure.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a schematic flow chart of a monitoring method according to a first embodiment of the invention;
FIG. 2 is a schematic diagram of dam crest crack distribution and measuring point arrangement according to a second embodiment of the present invention;
FIG. 3 is a schematic illustration of dam axis crack in accordance with a second embodiment of the present invention;
FIG. 4 is a schematic diagram of a probe pit inspection situation according to a second embodiment of the present invention;
FIG. 5 is a comparison of actual and predicted values of the displacement of a dam crest along a river in accordance with the second embodiment of the present invention;
FIG. 6 is a block diagram of a dam crest forward river displacement monitoring station residual sequence D-Vine in accordance with the second embodiment of the present invention;
FIG. 7 is a schematic diagram of a residual sequence frequency distribution and a fitting probability density function of a measurement point according to a second embodiment of the present invention;
FIG. 8 is a schematic diagram of damage monitoring results of a multi-station combined structure health monitoring method according to a second embodiment of the present invention;
fig. 9 is a schematic diagram of a monitoring interval of a downstream displacement of a top of a river at a measuring point TP16 according to a second embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
Referring to fig. 1, the embodiment provides a multi-measuring point joint monitoring method for dam structure damage, which includes the following steps:
step one: firstly, arranging monitoring points at different positions of a dam to be monitored to monitor the appearance deformation of the dam, then respectively fitting a corresponding prediction model for each measuring point according to monitoring data obtained by measuring the monitoring points arranged in the dam, and then calculating residual errors between actual measurement values of each monitoring point and prediction values of each prediction model to obtain each residual error sequence;
step two: because the residual sequences belong to time sequence variables and have a certain sequence, in the embodiment, kendall rank correlation coefficients are selected as measurement indexes for measuring the correlation of the residual sequences to analyze the correlation and the consistency among the residual sequences, and X { X } is set 1 ，x 2 ，L，x n Sum Y { Y } 1 ，y 2 ，L，y n The residual sequence of two measuring points (x) i ，y i ) And (x) j ，y j ) Residual data pairs respectively of points at two sides of i and j moments, if x i ＜x j And y is i ＜y j Or x i ＞x j And y is i ＞y j Then call (x) i ，y i ) And (x) j ，y j ) If x is consistent with i ＜x j And y is i ＞y j Or x i ＞x j And y is i ＜y j Then call (x) i ，y i ) And (x) j ，y j ) Non-uniform, when X, Y two sides are shared
And c represents a consistent residual function pair, d represents a non-consistent residual data pair, and Kendall rank correlation coefficient is defined as:
and then selecting a Vine structure with the maximum sum of Kendall rank correlation coefficient absolute values by using a maximum tree generation algorithm, and expressing the Vine structure as follows:
wherein T is i Is the firstAggregation of all possible tree structures in i trees, t i Representing the structure of a particular ith tree, e is tree t i Any edge, delta i.j Kendall rank correlation coefficients of a pair of residual sequences corresponding to the e-edge are represented;
fitting probability density function of residual sequence distribution by using kernel density estimation method
The expression is:
in the method, in the process of the invention,
to fit the probability density function, x i K is a Gaussian kernel function, h is bandwidth and is selected according to optimization of integral mean square error, and edge distribution of each residual sequence used for constructing a Vine Copula model is obtained;
step three: firstly, selecting a binary Copula function C (-) for each pair of edge distributions in the Vine structure in the second step according to a red pool information criterion (AIC), and selecting a Copula function type corresponding to the minimum AIC value, wherein the formula is as follows:
AIC＝2K-2ln(L)
wherein K is the number of parameters, for five Copula functions of Frank Copula, clayton Copula, gumbel Copula, gaussian Copula and t-Copula, K is 2 except for the t-Copula function, the other types are 1, and L is the maximum value of the likelihood function;
and estimating parameters of the binary Copula function selected by each edge in Vine by using a Maximum Likelihood Estimation (MLE), wherein the calculation formula of the likelihood function is as follows:
where n represents the number of samples, θ= (θ) 1 ，θ 2 ，…θ k ) The parameter vector representing the Vine Copula model, C (·) is the density function of the Copula function C (·), F (·) is the edge distribution, let
For maximum likelihood estimation of θ, there are
A set of likelihood equations according to:
the model parameters can be solved;
then constructing a VineCopula model according to the selected Copula function and the selected binary Copula function;
step four: firstly, calculating a difference value between currently acquired monitoring data and a predicted value of a prediction model, obtaining a current residual sequence, substituting the current residual sequence into a probability density function in the second step to obtain edge distribution, substituting the edge distribution into a VineCopula model in the third step after judging to obtain a corresponding combined Cumulative Distribution Function (CDF) value at each moment, wherein the edge cumulative distribution function value judging formula is as follows:
wherein CDF is the value of the cumulative distribution function of the edge;
step five: firstly determining that the abnormal threshold value is alpha/2 (alpha is the significance level, taking 1%), comparing the combined cumulative distribution function value obtained in the fourth step with the abnormal threshold value, if the combined cumulative distribution function value is lower than the abnormal threshold value, the structural state of the dam is abnormal, and if the combined cumulative distribution function value is higher than the abnormal threshold value, the structural state of the dam is normal, so that whether the dam structure is damaged or not is monitored.
Example two
Referring to fig. 2, 3, 4, 5, 6, 7, 8 and 9, in this embodiment, a gravelly earth core rockfill dam in southwest China is used as a monitoring object, the maximum dam height of the rockfill dam is 186m, the dam top height is 856.00m, the dam top width is 14m, the normal water storage level 850.00m of a reservoir, the flood season operation limit water level 841.00m, the dead water level 790.00m and the day 11 in 2020 are found to have an open crack near the dam top axis when the rockfill dam is routinely inspected, and the width is about 2mm, as shown in fig. 2 and 3. In order to ascertain the crack depth, pit inspection was performed on site on days 3 to 5 of 11 in 2020, as shown in fig. 4. Pit inspection showed that the maximum depth of the crack did not exceed 1.5m and did not extend to the core wall. In order to prevent rainwater from penetrating along cracks to further influence dam seepage stability and damage of structural integrity possibly caused by crack continuous development, the management unit adopts fine sand, asphalt and other materials to seal the longitudinal cracks of the dam top axis. After repair, no new open cracks were found by 31 days 12 to 2020.
The engineering adopts an intelligent measuring robot to monitor the appearance deformation of the rock-fill dam. In order to verify the effectiveness of the method for monitoring the structural damage, 5 monitoring points TP12, TP13, TP14, TP15 and TP16 near the dam crest crack are selected as analysis objects, as shown in figure 2, and the forward river displacement monitoring data from 10 days in 2018 to 31 days in 12 and 2020 are analyzed. The monitoring data is divided into two segments in time. The fitting section is from 8 months of 2018 to 10 months of 2020, the inspection section is from 11 months of 2020 to 31 months of 2020, and the corresponding monitoring data are respectively called a data set 1 and a data set 2.
The method of the invention is adopted to monitor the structural health of the earth-rock dam, and comprises the following specific steps:
s1: the method comprises the steps of obtaining a prediction model of the forward river displacement of the dam tops of all measuring points based on a data set 1, wherein the earth-rock dam displacement mainly comprises a water pressure component, a temperature component and an aging component, wherein the aging component is related to the rheological deformation of a rock-fill body, and constructing the prediction model by adopting an aging separation method in order to reflect the deformation which is unrecoverable with time, so that the expression of the prediction model is as follows:
in the method, in the process of the invention,
for the expression of the hydraulic pressure component in the river displacement, H is the reservoir water level,
for the expression of the temperature component in the forward displacement, θ= (t-t) 0 )/100，t 0 For monitoring the start date, t is the monitoring date, < >>
As expression of aging component, a i 、b 1 、b 2 、c 1 、c 2 、c 3 、c 4 As model coefficients, d is a constant term, according to the monitoring data in the data set 1, obtaining predicted model coefficients as shown in the following table 1, and comparing the monitoring data of the dam body along river displacement with the predicted model fitting values as shown in fig. 5;
TABLE 1 predictive model coefficient table
| Measuring point | a 1 | a 2 | b 1 | b 2 | c 1 | c 2 | c 3 | c 4 | d |
| TP12 | -16.83 | 0.01 | -3.04 | -6.63 | -11.58 | -2.36E-28 | 17.96 | -49.19 | 6557.58 |
| TP13 | -19.97 | 0.01 | -5.99 | -8.27 | 7132.65 | -2.52E+01 | -2.29 | -6976.4 | 7794.23 |
| TP14 | -21.61 | 0.01 | -5.82 | -8.40 | 65.63 | -2.38E+09 | -6.15 | 85.90 | 8463.77 |
| TP15 | -21.65 | 0.01 | -4.87 | -7.99 | 68.66 | -3.79E+07 | -5.04 | 33.53 | 8519.68 |
| TP16 | -18.31 | 0.01 | -3.02 | -6.16 | 27.98 | -1.42E+09 | -6.03 | 124.82 | 7253.51 |
S2: the difference between the dam crest forward monitoring data and the fitting value of the prediction model is used as a residual sequence corresponding to each measuring point, the calculation result of the dependency measurement index Kendall rank correlation coefficient between the residual sequences corresponding to each measuring point is shown in table 2, wherein the numerical value of the rank correlation coefficient of the thickened italics is larger, as can be known from table 2, the Kendall rank correlation coefficient between the residual sequences of each measuring point is more than 0.55, the correlation is certain, the rank correlation coefficient between each pair of adjacent measuring point residual sequences is the largest, namely, the correlation between the residual sequences of the measuring points TP12 and TP13, TP13 and TP14, TP14 and TP15 and TP16 is the strongest, and the overall characteristic of serial connection is realized, so the D-Vine structure shown in figure 6 is adopted;
TABLE 2 prediction model coefficient table
| TP12 | TP13 | TP14 | TP15 | TP16 | |
| TP12 | 1.0000 | 0.7230 | 0.7122 | 0.6566 | 0.5578 |
| TP13 | 0.7230 | 1.0000 | 0.8182 | 0.7976 | 0.6588 |
| TP14 | 0.7122 | 0.8182 | 1.0000 | 0.8472 | 0.6987 |
| TP15 | 0.6566 | 0.7976 | 0.8472 | 1.0000 | 0.7230 |
| TP16 | 0.5578 | 0.6588 | 0.6987 | 0.7230 | 1.0000 |
S3: fitting residual sequence distribution by a kernel density estimation method to be used as edge distribution required by construction of a Vine Copula model, wherein the kernel density estimation bandwidths of the obtained residual sequences corresponding to the measuring points TP12 to TP16 are 0.5182, 0.6290, 0.7354, 0.7677 and 0.7243 respectively, and the fitted probability density curve and the frequency distribution of the residual sequence are shown in figure 7;
s4: selecting the optimal Copula connection function type of each connection edge in the D-Vine by using the red pool information criterion, and determining the parameters of each binary Copula function in the Vine Copula model by using a maximum likelihood estimation method, wherein the parameters are shown in Table 3;
TABLE 3 Copula function type selection and parameter Table in Vine-Copula model
S5: based on the data set 1 and the prediction model constructed in the step S1, a new residual sequence is obtained, new edge distribution is obtained according to the kernel density function in the step S3, then the new edge distribution is substituted into the constructed Vine-Copula model to be calculated to obtain a corresponding combined CDF value, and finally the result of structural damage monitoring obtained by the multi-measuring-point combined structural health monitoring method is shown in a graph 8, wherein a horizontal dotted line represents an abnormal threshold value of CDF of 0.005 (alpha=1%), a left vertical dotted line represents a time point of finding a crack, and a right vertical dotted line represents a time point of performing plugging treatment, and as can be seen from the graph 8, the combined monitoring method sequentially emits 19 alarms within 3 periods of 9 months 12 to 9 months 21 days, 10 months 9 to 10 months 28 days, and 11 months 3 days to 11 months 7 days;
s6: in order to compare the effectiveness and accuracy of the method provided by the invention and the method based on the single-measuring-point monitoring model on anomaly monitoring, the residual sequences of the measuring points obtained in the step S2 are shown in a table 4 by adopting a confidence interval estimation method to determine left and right side split values of the residual when the significance level is 1%, and then the operation warning value of each measuring point is determined according to the prediction model constructed in the step S1, wherein the monitoring result shows that in the period of 11 days to 31 days of 2020 8 months and 11 months, when the traditional single-measuring-point monitoring method is adopted, only the monitoring data of TP16 in 5 measuring points exceeds a monitoring interval, and the number of alarm times in three months is very small and only 5 times are shown in fig. 9.
Table 4 monitoring station residual sequence left and right side quantile values (α=99%)
| TP12 | TP13 | TP14 | TP15 | TP16 | |
| Right side quantile value | 4.414 | 5.198 | 7.035 | 5.546 | 4.496 |
| Left-hand quantile value | -4.089 | -5.539 | -5.882 | -5.720 | -4.372 |
The method comprises the steps of constructing a prediction model of forward displacement of a dam top of each measuring point by using monitoring data in the period from 8 th month 10 of a earth-rock dam 2018 to 10 th month 8 of a 2020, fitting probability distribution of residual sequences corresponding to each measuring point, and constructing a vine copula model for calculating multidimensional joint probability distribution. And then defining the monitoring data from 11 days in 2020 to 31 days in 2020 as newly acquired dam crest along river displacement monitoring data, and carrying out structural damage monitoring. And simultaneously, the traditional single-measuring-point monitoring method is used for monitoring the abnormal condition of the monitoring data, so that the effectiveness and the accuracy of the method are tested.
As can be seen from the results shown in fig. 8 and 9, the number of alarms in the combined monitoring method is far greater than that in the monitoring method based on the single-station monitoring model, and the generated alarms are also more dense in time distribution. In the security management of a dam, this situation cannot be considered as simply a data anomaly, but must be considered as an anomaly caused by structural damage. The damage monitoring results show that the monitoring data of the measuring points at the top of the dam from 9 months and 12 days show that the abnormal condition gives an alarm, so that it can be presumed that the crack at the top of the dam is in a gradually developing state before the inspection is found at 11 months and 2 days. In addition, after the cracks were plugged at 11 months and 9 days, no new development of the dam top cracks occurred by 12 months and 31 days, presumably because the reservoir water level change trend was changed from rising to gradually falling during this period. Accordingly, no condition occurs during which the joint CDF value exceeds the anomaly threshold. This illustrates that the method proposed herein can effectively and accurately monitor dam structure damage compared to monitoring methods based on single-site monitoring models.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (8)
1. The multi-measuring point joint monitoring method for dam structure damage is characterized by comprising the following steps of:
step one: firstly, arranging monitoring points at different positions of a dam to be monitored to monitor the appearance deformation of the dam, then respectively fitting a corresponding prediction model for each measuring point according to monitoring data obtained by measuring the monitoring points arranged in the dam, and then calculating residual errors between actual measurement values of each monitoring point and prediction values of each prediction model to obtain each residual error sequence;
step two: firstly, selecting Kendall rank correlation coefficients as a measurement index for measuring residual sequence dependence to analyze the correlation and consistency among residual sequences, then selecting a Vine structure with the largest sum of Kendall rank correlation coefficient absolute values by using a maximum tree generation algorithm, and fitting a probability density function of residual sequence distribution by using a kernel density estimation method to obtain edge distribution of each residual sequence;
step three: firstly, distributing and selecting a binary Copula function for each pair of edges in a Vine structure in the second step according to a red pool information criterion, estimating parameters of the binary Copula function selected for each edge in Vine by using a maximum likelihood estimation method, and constructing a VineCoula model according to the selected Copula function and the selected binary Copula function;
step four: firstly, calculating the difference between the currently acquired monitoring data and the predicted value of the predicted model, obtaining a current residual sequence, substituting the current residual sequence into the probability density function in the second step to obtain edge distribution, and substituting the edge distribution into the Vine Copula model in the third step after judging to obtain a corresponding combined cumulative distribution function value at each moment;
step five: firstly determining an abnormal threshold value, then comparing the combined cumulative distribution function value obtained in the fourth step with the abnormal threshold value, and judging whether damage exists in the dam structure according to a comparison result.
2. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the second step, when the correlation and consistency between residual sequences are analyzed, X { X }, is set 1 ，x 2 ，L，x n Sum Y { Y } 1 ，y 2 ，L，y n The residual sequence of two measuring points (x) i ，y i ) And (x) j ，y j ) Residual data pairs respectively of points at two sides of i and j moments, if x i ＜x j And y is i ＜y j Or x i ＞x j And y is i ＞y j Then call (x) i ，y i ) And (x) j ，y j ) If x is consistent with i ＜x j And y is i ＞y j Or x i ＞x j And y is i ＜y j Then call (x) i ，y i ) And (x) j ，y j ) Non-uniform, when X, Y two sides are shared
A different residual data pair, kendall rank correlation coefficient, is defined as:
where c represents a consistent residual function pair and d represents a non-consistent residual data pair.
3. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the second step, the formula of the maximum tree generation algorithm is as follows:
wherein T is i T is the set of all possible tree structures in the ith tree i Representing the structure of a particular ith tree, e is tree t i Any edge, delta i.j And Kendall rank correlation coefficients of a pair of residual sequences corresponding to the e-side are represented.
4. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the second step, the probability density function has the expression:
in the method, in the process of the invention,
to fit the probability density function, x i K is a Gaussian kernel function, h is a bandwidth and is selected according to the optimization of the integral mean square error. />
5. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the third step, the formula for calculating the erythro pool information criterion is as follows:
AIC＝2K-2ln(L)
where K is the number of parameters, K is 2 for the five Copula functions of FrankCopula, claytonCopula, gumbelCopula, gaussianCopula, t-Copula, the remaining types are 1, and L is the maximum value of the likelihood function.
6. The method for multi-measuring point joint monitoring of dam structure damage according to claim 5, wherein the method comprises the following steps: in the third step, the calculation formula of the likelihood function is as follows:
where n represents the number of samples, θ= (θ) 1 ，θ 2 ，…θ k ) The parameter vector representing the Vine Copula model, C (·) is the density function of the Copula function C (·), and F (·) is the edge distribution.
7. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the fourth step, the edge cumulative distribution function value judging formula is:
where CDF is the value of the edge cumulative distribution function.
8. The method for multi-measuring point joint monitoring of dam structure damage according to claim 1, wherein the method comprises the following steps: in the fifth step, when the combined cumulative distribution function value is compared with the abnormal threshold value, if the combined cumulative distribution function value is lower than the abnormal threshold value, the dam structure is abnormal, and if the combined cumulative distribution function value is higher than the abnormal threshold value, the dam structure is normal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310211580.8A CN116167240A (en) | 2023-03-07 | 2023-03-07 | Multi-measuring-point combined monitoring method for dam structure damage |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310211580.8A CN116167240A (en) | 2023-03-07 | 2023-03-07 | Multi-measuring-point combined monitoring method for dam structure damage |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116167240A true CN116167240A (en) | 2023-05-26 |
Family
ID=86416344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310211580.8A Pending CN116167240A (en) | 2023-03-07 | 2023-03-07 | Multi-measuring-point combined monitoring method for dam structure damage |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116167240A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736781A (en) * | 2023-08-15 | 2023-09-12 | 国网浙江省电力有限公司杭州供电公司 | Safety state monitoring method and device for industrial automation control equipment |
-
2023
- 2023-03-07 CN CN202310211580.8A patent/CN116167240A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736781A (en) * | 2023-08-15 | 2023-09-12 | 国网浙江省电力有限公司杭州供电公司 | Safety state monitoring method and device for industrial automation control equipment |
| CN116736781B (en) * | 2023-08-15 | 2023-11-03 | 国网浙江省电力有限公司杭州供电公司 | Safety state monitoring method and device for industrial automation control equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Su et al. | Time-varying identification model for dam behavior considering structural reinforcement | |
| CN111401529B (en) | River ecological flow supervision method based on remote sensing technology | |
| CN111695177B (en) | Dam mechanics parameter random inversion method and system based on displacement field monitoring data | |
| Muzzammil et al. | An optimization technique for estimation of rating curve parameters | |
| CN116167240A (en) | Multi-measuring-point combined monitoring method for dam structure damage | |
| CN109002888A (en) | A kind of red tide prewarning method | |
| CN112528563B (en) | Urban waterlogging early warning method based on SVM algorithm | |
| CN115774953A (en) | Pollution space-time risk supervision and evaluation system and method based on data processing | |
| CN111340284A (en) | Intelligent waterwheel room water level prediction method based on long-time memory network | |
| CN110851897B (en) | Aqueduct stress-strain prediction method under multi-factor association | |
| CN114896459A (en) | Rainwater pipe network sewage steal-drainage tracing method utilizing directed graph | |
| CN113128098B (en) | Concrete dam singular value detection method and equipment based on DE-LOF | |
| CN114021078A (en) | Optimal statistical model optimization method for dam monitoring quantity | |
| CN114548680A (en) | Method and system for automatically calibrating parameters of urban storm flood management model | |
| CN113887119B (en) | River water quality prediction method based on SARIMA-LSTM | |
| Rohaimi et al. | 3 Hours ahead of time flood water level prediction using NNARX structure: Case study pahang | |
| Jiang et al. | Method of fusion diagnosis for dam service status based on joint distribution function of multiple points | |
| CN110847112B (en) | River flood discharge early warning method based on hydraulics simulation | |
| CN114781156B (en) | BIM-based channel maintenance measure analysis method | |
| CN115713312A (en) | Combined identification method and device for unfavorable working conditions of hydraulic and hydroelectric engineering environmental quantity | |
| CN113807709B (en) | Multi-target lake area water safety evaluation method based on water condition elements | |
| CN115310258A (en) | Underground pipeline service life prediction method based on LSTM | |
| Lyu et al. | Water level prediction model based on GCN and LSTM | |
| CN112632838A (en) | Dynamic evaluation and early warning method for cracking risk of dam crest of high core wall rock-fill dam | |
| CN117709488B (en) | Dam seepage prediction method based on RUN-XGBoost |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |