A rectangular caisson structure stability prediction method based on geometric-damage coupling mechanism

Abstract

The application discloses a rectangular caisson structure stability prediction method based on a geometry-damage coupling mechanism, fuses displacement, strain and microseismic multi-source monitoring data, extracts multi-dimensional geometry parameters such as caisson settlement difference, overall inclination angle, wall plate bulging deformation rate and diagonal shear deformation rate, and constructs a geometry stability characteristic vector; a damage variable and stiffness degradation coefficient quantitative model is established, a neural network embedding a damage evolution physical constraint is used to construct a coupling mapping relationship, a geometry-damage-stiffness degradation unified evaluation system is formed, and combined with stage adaptive weight adjustment and time sequence recursive algorithm, quantitative prediction and risk classification of the caisson full life cycle stability are realized. The method is suitable for urban power grid underground caisson engineering, can effectively improve prediction accuracy and early warning timeliness, and solves the technical problem that geometry deformation and internal damage analysis are mutually disjointed.

Description

Technical Field

[0001] This invention relates to the field of underground engineering structure safety monitoring and stability assessment technology, and in particular to a method for predicting the stability of rectangular caisson structures based on a geometry-damage coupling mechanism. This method is applicable to the full life cycle stability prediction and risk assessment of rectangular caisson structures such as urban power grid underground substations and cable shafts. Background Technology

[0002] Caisson structures, with their advantages of small construction footprint, good structural integrity, and low impact on the surrounding environment, are widely used in underground engineering projects such as urban power grid substations and cable shafts. The entire life cycle of a caisson structure encompasses multiple stages, including sinking construction, bottom sealing, backfilling, and long-term operation. Due to the coupled effects of multiple factors such as geological heterogeneity, groundwater level fluctuations, construction disturbances, changes in surrounding loads, and long-term material deterioration, structural deformation and internal damage exhibit complex time-varying evolution characteristics. If structural stability is out of control, it can easily lead to engineering accidents such as overall tilting, wall panel cracking, uneven settlement, or even overall instability, seriously threatening the safe operation of underground power grid facilities. Therefore, achieving high-precision stability prediction and risk warning throughout the entire process of caisson structures is of significant engineering value in ensuring the long-term safety of underground engineering projects.

[0003] Existing technologies for monitoring and assessing the stability of caisson structures mainly fall into two categories: one focuses on monitoring external geometric morphology, acquiring geometric parameters such as overall settlement and tilt of the caisson by setting up displacement measuring points and using three-dimensional laser scanning, thereby evaluating the overall deformation state of the structure. This type of method offers convenient data acquisition and broad coverage, but it only reflects the apparent deformation characteristics of the structure and cannot reveal the internal stress redistribution and material damage evolution process. When internal damage has not yet caused significant geometric deformation, it is difficult to identify early potential risks, resulting in a lag in early warning.

[0004] Another type of approach focuses on monitoring the internal mechanical state of the structure. It uses distributed optical fibers, strain sensors, and microseismic monitoring devices to acquire information such as internal strain and energy release from microcracks, which is then used to assess the degree of local damage. While this approach can capture the internal mechanical response of the structure, it has limited monitoring points and lacks a quantitative correlation with overall geometric deformation. Therefore, it cannot explain the correspondence between geometric anomalies and internal damage, making it difficult to support a systematic evaluation of the overall structural stability.

[0005] In addition, some existing studies have introduced data-driven algorithms to predict structural state, but most of them treat geometric monitoring data and mechanical monitoring data as independent input variables without embedding a clear physical evolution mechanism. They cannot construct a coupled mapping relationship between geometric morphology evolution and internal damage degradation, resulting in insufficient physical interpretability of the prediction results. It is difficult to achieve quantitative extrapolation from apparent geometric anomalies to structural instability risk, and cannot meet the needs of dynamic assessment of the stability of caissons throughout their entire life cycle.

[0006] In summary, existing technologies generally have the following shortcomings: First, there is a lack of a unified quantitative expression system for the evolution of the external geometric shape and the degradation of internal damage in caissons, and the two are isolated from each other; second, geometric monitoring and mechanical monitoring have not formed a linkage mechanism, making it difficult to support a comprehensive evaluation of overall stability; and third, there is a lack of coupled deduction models with physical constraints, making it impossible to predict the stability evolution of caisson structures throughout their entire life cycle and provide early warning of risks. Summary of the Invention

[0007] To address the aforementioned deficiencies in existing technologies, the present invention aims to provide a method for predicting the stability of rectangular caisson structures based on a geometry-damage coupling mechanism. By constructing a nonlinear coupling mapping relationship between the caisson's geometric evolution parameters and internal damage variables, an integrated evaluation framework for geometry-damage-stiffness degradation is established. This enables quantitative extrapolation of the entire process of the caisson structure, from apparent deformation to internal damage and then to stability degradation. Furthermore, by combining stage-adaptive weighting and temporal evolution prediction, the accuracy of caisson life-cycle stability prediction and the timeliness of risk warning are improved.

[0008] Specifically, this method provides a stability prediction method based on a geometry-damage coupling mechanism for rectangular caisson structures in urban power grid underground engineering. Addressing the stability requirements of rectangular caisson structures such as underground substations and cable shafts in State Grid under complex urban environments and long-term operating conditions, this method constructs a nonlinear coupling mapping relationship between the surface geometric evolution parameters of the rectangular caisson structure and internal damage variables. This enables quantitative deduction from external geometric anomalies to internal damage evolution and mechanical performance degradation. First, three-dimensional displacement field data of the rectangular caisson structure are acquired, and the overall settlement, uneven settlement coefficient, overall tilt angle, deformation rate of the long side wall, deformation rate of the short side wall, diagonal shear deformation rate, and corner... This invention establishes a geometric morphological characteristic parameter system for rectangular caissons by using geometric stability indices such as the deformation coefficient. Simultaneously, it collects internal strain field, temperature field, and microseismic energy release information to construct a model of structural damage variables and stiffness degradation coefficients. Based on this, a coupling mapping function between geometric parameters and damage variables is established, forming a unified expression system of geometry-damage-stiffness degradation. Furthermore, a stage-adaptive stability evaluation model is constructed to dynamically adjust the weights of geometric characteristic parameters and damage variables at different construction stages. Finally, through time-series evolution prediction, the life-cycle stability state and risk level of the rectangular caisson structure are obtained, enabling prediction of the entire process from external geometric evolution to internal damage development. This invention overcomes the technical limitations of independent analysis of geometric deformation and internal damage in traditional rectangular caisson stability monitoring, establishing a coupled prediction mechanism between geometric evolution and damage degradation. This improves the accuracy and reliability of rectangular caisson structure stability prediction, providing technical support for the long-term safe operation of rectangular caisson structures in underground engineering.

[0009] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0010] A method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism includes the following steps:

[0011] Step 1: Monitoring Data Acquisition and Geometric Stability Feature Vector Construction

[0012] High-precision displacement monitoring points and three-dimensional laser scanning equipment were deployed on the outer wall and top of the rectangular caisson to collect displacement monitoring data and three-dimensional point cloud data during the entire process of caisson sinking, bottom sealing, backfilling and operation. After the original monitoring data underwent coordinate transformation, gross error removal and smoothing and noise reduction preprocessing, multi-dimensional geometric deformation parameters were extracted to construct the geometric stability feature vector of the caisson.

[0013] Furthermore, the multi-dimensional geometric deformation parameters include: the overall settlement of the caisson. Overall tilt angle 1. Long side panel bulging deformation rate 1. Short side panel bulging deformation rate Diagonal shear deformation rate and the four-corner coordination deformation coefficient The calculation methods for each parameter are as follows:

[0014] (1) Calculate the differential settlement and the coefficient of uneven settlement:

[0015] Set time The maximum settlement, minimum settlement, and average settlement are respectively , and Then the differential settlement and the coefficient of uneven settlement are respectively:

[0016]

[0017]

[0018] (2) Overall tilt angle Let the average settlement of the top of the four well walls on the east, west, south, and north sides be respectively... , , and Based on the calculations of the caisson length L and width B, the overall tilt angle is calculated using the following formula:

[0019]

[0020] (3) bulging deformation rate of long side panel 1. Short side panel bulging deformation rate The maximum normal bulging displacement of the long side panel was obtained by fitting the 3D point cloud. and the maximum normal bulge displacement of the short side panel The bulging deformation rates of the long side panel and the short side panel are respectively:

[0021]

[0022] In the formula, These represent the maximum normal bulge displacements of the long and short side panels, respectively.

[0023] (4) Diagonal shear deformation rate The result is calculated based on the difference in length changes of the two diagonals of the caisson and the initial average diagonal length.

[0024]

[0025] In the formula, These represent the changes in the lengths of the two diagonals, The average length of the two diagonals at the initial moment;

[0026] (5) Coefficient of four-corner compatibility : The ratio of the range of settlement at the four corners of the caisson to the average settlement, expressed as a percentage.

[0027]

[0028] Let be the settlement at the j-th corner point, where , This represents the average settlement value of the caisson.

[0029] Construct the caisson geometric stability feature vector based on the above parameters. :

[0030] .

[0031] Step 2: Modeling structural damage variables and stiffness degradation coefficients

[0032] An array of fiber optic strain sensors, temperature compensation units, and microseismic monitoring units were deployed at key sections inside the caisson structure to collect data on the strain field ε(x,y,z,t), temperature field T(x,y,z,t), and microseismic energy release E within the caisson. m (t) Data; Based on the monitoring data, the current equivalent elastic modulus and overall structural stiffness are inverted, and a quantitative relationship model between structural damage variables and stiffness degradation coefficient is constructed.

[0033] Furthermore, temperature compensation is first applied to the total strain measured by the optical fiber to calculate the mechanical strain:

[0034]

[0035] In the formula, For the measured total strain, Coefficient of linear expansion of materials This represents the change in temperature.

[0036] Equivalent stress increment based on key sections Equivalent effect change increment The equivalent elastic modulus at time t is obtained by inversion. :

[0037]

[0038] Define structural damage variables Characterizes the degree of material performance degradation:

[0039]

[0040] In the formula, The initial equivalent elastic modulus of the caisson structure is determined through benchmark monitoring before construction.

[0041] The overall structural stiffness at time t is calculated by the ratio of the representative load increment to the comprehensive displacement increment. :

[0042]

[0043] And construct the stiffness degradation coefficient Linear relationship with damage variables:

[0044]

[0045] In the formula, The initial overall stiffness of the caisson. The damage influence coefficient; By using multiple sets of synchronously monitored damage variables and stiffness data, an objective function was constructed and calibrated using the least squares method.

[0046]

[0047] In the formula, To determine the number of calibrated sample groups, Let the damage variable be the sample group k. Let be the measured overall stiffness of the k-th sample group.

[0048] Step 3: Construction of the geometry-damage coupling mapping model

[0049] Using geometric stability eigenvectors as input and structural damage variables as output, a nonlinear coupled mapping model with embedded physical constraints of damage evolution is constructed to establish a quantitative correlation between external geometric deformation and internal damage evolution.

[0050] Furthermore, the coupling mapping model is implemented using a physically constrained neural network. The network structure includes an input layer, a hidden layer, and an output layer: the input layer nodes correspond to normalized geometric feature parameters, the hidden layer uses the hyperbolic tangent activation function, and the output layer uses the sigmoid activation function, with the output being the predicted damage variable; the mapping relationship is expressed as:

[0051]

[0052] In the formula, For neural network mapping functions, To predict damage variables.

[0053] The training loss function of the neural network is composed of a weighted sum of a data fitting term and a physical residual term, and its expression is:

[0054]

[0055] In the formula, the first term is the data fitting term, which characterizes the deviation between the predicted damage and the measured damage; the second term is the physical residual term, which characterizes the deviation between the damage evolution rate and the damage constitutive equation. These are the weighting coefficients; Let be the damage evolution function, and its expression is:

[0056]

[0057] In the formula, For equivalent stress, The initial damage stress; For equivalent change, This represents the initial damage strain; For micro-vibration energy, For reference microseismic energy; , , The damage evolution parameters to be identified are obtained by using the least squares method based on monitoring data.

[0058] Step 4: Unified expression of geometry-damage-stiffness degradation and calculation of stability index

[0059] Based on the predicted damage variables output by the coupled mapping model, and combined with the initial overall stiffness and critical instability stiffness, a unified expression system for geometry-damage-stiffness degradation is constructed to calculate the stability index of the caisson at the current moment.

[0060] Furthermore, the current overall stiffness is derived from the predicted damage variables. :

[0061]

[0062] Define stability index The current overall stiffness and critical instability stiffness The ratio:

[0063]

[0064] In the formula, the critical buckling stiffness is... The limit state was determined through structural design verification and three-dimensional finite element analysis of the caisson-soil mass.

[0065] Step 5: Stage Adaptive Weight Adjustment and Comprehensive Evaluation

[0066] For different stages of caisson sinking, bottom sealing, backfilling and operation, a single-factor sensitivity analysis method is used to determine the adaptive weights of each geometric deformation parameter and damage variable at the corresponding stage, and generate stage adaptive comprehensive evaluation quantity.

[0067] The specific process is as follows: combine all geometric deformation parameters and damage variables into a stage evaluation vector. Within each stage, a set proportion of single-factor perturbation is applied to each feature parameter. The change in the stability index is calculated using a calibrated coupled mapping model, and this change is used as the sensitivity of the corresponding feature. The sensitivity of each feature within the same stage is normalized to obtain the adaptive weights corresponding to each feature in that stage. Furthermore, the sum of all weights within the same stage is equal to 1; finally, the normalized feature values ​​are summed with their corresponding weights to obtain the adaptive comprehensive evaluation value for that stage. :

[0068]

[0069] In the formula, For geometric or damage characteristic parameters, it is the first The normalized values ​​of each feature parameter. ω represents the total number of characteristic parameters. i (t) is the weighting coefficient, which is automatically updated and adjusted according to the construction stage and environmental load changes, thereby enhancing the stage sensitivity of stability evaluation.

[0070] Step 6: Time-series evolution prediction and risk level classification

[0071] Based on the current monitoring data, the damage evolution rate is identified, and the damage variables and stability indicators at future moments are predicted by a first-order time-series recursive algorithm. The stability risk level of the caisson structure is then classified according to a preset threshold.

[0072] Furthermore, in the future The predicted damage variable at time step is calculated using a first-order recursive formula:

[0073]

[0074] In the formula, To predict the time step, This represents the damage evolution rate at the current moment.

[0075] The predictive stability index corresponding to future time points is:

[0076]

[0077] Based on the stability index values, the stability state of the caisson structure is divided into three levels:

[0078] Safe status:

[0079] When S_f≥1.2, the structure has sufficient stiffness reserve and good stability;

[0080] Warning status:

[0081] When 1.0 ≤ S_f < 1.2, the structural stiffness reserve decreases, posing a potential risk of instability, and monitoring needs to be strengthened.

[0082] Instability risk status:

[0083] When S_f < 1.0, the structural stiffness is below the critical value, the risk of instability is high, and timely control measures need to be taken.

[0084] An electronic device, the device comprising:

[0085] At least one processor; and

[0086] A memory communicatively connected to the at least one processor; wherein,

[0087] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the above-described method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism.

[0088] A computer-readable storage medium storing computer instructions for causing a processor to execute the above-described method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism.

[0089] Compared with the prior art, the present invention has the following beneficial effects:

[0090] 1) A unified quantitative system for geometry-damage-stiffness degradation was constructed, overcoming the technical limitations of traditional monitoring methods that separate geometric deformation from internal damage. Through the coupled mapping of multi-dimensional geometric features and damage variables, quantitative deduction from apparent geometric anomalies to internal damage degradation and then to overall stiffness decay was achieved, solving the problem that existing methods cannot correlate geometric deformation with mechanical state;

[0091] 2) A coupled mapping model is constructed using a neural network with embedded physical constraints on damage evolution. This not only leverages the nonlinear fitting capability of data-driven methods but also improves the physical interpretability and accuracy of the prediction results by constraining the damage evolution law through physical residual terms. Validation shows that the average absolute error of damage prediction using this method is reduced by more than 40% compared to neural networks without physical constraints and by more than 70% compared to linear regression methods. The accuracy of grade identification can reach more than 96%.

[0092] 3) A phase-adaptive weight adjustment mechanism is proposed to dynamically adjust the characteristic weights for the risk-dominant factors in different construction and operation phases of the caisson. In the sinking phase, the focus is on geometric parameters such as settlement and tilt, while in the operation phase, the focus is on damage variables and coordinated deformation coefficients, thereby enhancing the phase-specificity and sensitivity of stability evaluation.

[0093] 4) A time-series recursive evolution prediction model has been established, which can predict future damage development and stability change trends based on current monitoring data. It can provide early warning before stability indicators cross the risk threshold, providing decision support for the safety management of the entire life cycle of the caisson and effectively improving the initiative in risk disposal. Attached Figure Description

[0094] The accompanying drawings, as part of this invention, are provided to further illustrate the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation thereof. Clearly, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0095] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0096] Figure 2 This is a schematic diagram of the layout of the caisson structure monitoring device of the present invention;

[0097] Figure 3 A table of geometric monitoring data for a typical phase;

[0098] Figure 4 Table of damage variables and stiffness calibration data;

[0099] Figure 5 Table of prediction results for typical stages;

[0100] Figure 6 An adaptive weight table for different stages;

[0101] Figure 7 Table of validation results for different methods;

[0102] Figure 8 This is a schematic diagram of the electronic device used in the present invention, which is based on the geometry-damage coupling mechanism for predicting the stability of rectangular caisson structures. Detailed Implementation

[0103] To enable those skilled in the art to better understand the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention. It should be understood that the various steps described in the method embodiments of the present invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0104] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0105] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, any variations of the terms "comprising" and "having," etc., are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0106] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0107] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0108] Example 1

[0109] This embodiment uses a rectangular caisson in an underground power grid project as an example. The caisson has an outer length of 18m, a width of 14m, a total depth of 24m, a wall thickness of 1.0m, and a concrete strength grade of C40. The entire construction and operation process is divided into the sinking stage, the bottom sealing stage, the backfilling stage, and the operation stage. The stability prediction of this caisson is carried out using the method of this invention. The specific implementation process is as follows:

[0110] Step 1: Monitoring Data Acquisition and Geometric Feature Vector Construction

[0111] Monitoring data acquisition

[0112] according to Figure 2 As shown, 32 high-precision displacement monitoring points are set up on the outer wall and top of the caisson, and a three-dimensional laser scanning device is used to obtain the point cloud of the caisson wall surface; 16 electrical resistivity electrodes and 6 microseismic detectors are set up on the left and right sides of the caisson wall respectively; array-type optical fibers are set up at different elevations on the inner side of the caisson wall and temperature compensation units are configured; and three-component microseismic sensors and displacement sensors are set up at the bottom of the caisson.

[0113] The monitoring data acquisition frequency is set as follows: displacement data is acquired once every 1 hour, fiber optic strain and temperature data are acquired once every 1 minute, microseismic data is acquired continuously, and 3D point cloud data is acquired once every 3 days after the end of each construction stage and during the operation period.

[0114] Using the initial stability monitoring results completed before the caisson sealing construction as the baseline state, the initial equivalent elastic modulus Take 32.50 GPa as the initial overall stiffness. Pick N / m. Based on structural design calculations and three-dimensional finite element limit state analysis of the caisson-soil mass, the critical buckling stiffness is... Pick N / m.

[0115] Geometric feature vector construction

[0116] Unified coordinate transformation, gross error removal, and time synchronization are performed on displacement measurement points and 3D point cloud data. Gross error removal adopts the three-times-standard-deviation criterion, and continuous data is smoothed using a five-point moving average.

[0117] Set time The maximum settlement, minimum settlement, and average settlement are respectively , and Then the differential settlement and the coefficient of uneven settlement are respectively:

[0118]

[0119]

[0120] Let the average settlement of the top of the four well walls on the east, west, south, and north sides be respectively... , , and The length and width of the caisson are respectively and The overall tilt angle is then calculated using the following formula:

[0121]

[0122] The maximum normal bulge displacement of the long side panel was obtained by fitting the 3D point cloud. and the maximum normal bulge displacement of the short side panel The bulging deformation rates of the long side panel and the short side panel are respectively:

[0123]

[0124] Let the average length of the two diagonals at the initial time be... ,time The length changes are respectively and The diagonal shear deformation rate is:

[0125]

[0126] Let the settlement at the four corner points be... ,in Then the four corner coordination deformation coefficients are:

[0127]

[0128] This yields a caisson geometric stability feature vector consistent with the invention:

[0129]

[0130] This embodiment selects four typical stages, and the geometric monitoring results are shown in [the table below]. Figure 3 As shown.

[0131] Step 2: Damage Variable Calculation and Stiffness Parameter Calibration

[0132] The total strain measured by the arrayed fiber optic cable is first compensated for by temperature. Let the measured total strain be... The temperature change is The coefficient of linear expansion of the material is Then the mechanical strain is:

[0133]

[0134] Under the same load increment, based on the equivalent stress increment of the critical section. Equivalent effect change increment Invert the current equivalent elastic modulus:

[0135]

[0136] To ensure that the damage variable increases monotonically as the equivalent elastic modulus decreases, this embodiment defines the damage variable as follows:

[0137]

[0138] The overall stiffness is determined by the ratio of the representative load increment to the comprehensive displacement increment, i.e. The relationship between the damage variable and the stiffness degradation coefficient is as follows:

[0139]

[0140] in, This is the damage impact coefficient; to avoid giving it based on experience. Least square calibration was performed using synchronous monitoring data; for Given a set of samples, construct the objective function:

[0141]

[0142] right Differentiate and set it equal to zero:

[0143]

[0144] The calibration formula for the damage influence coefficient is obtained as follows:

[0145]

[0146] The inversion of four typical stages yielded and They are respectively Figure 4 As shown; substituting into the above formula, we get .

[0147] Step 3: Training the geometry-damage coupling mapping model

[0148] The components of the geometric feature vector are normalized according to their design control values ​​to obtain a dimensionless input vector. The physically constrained neural network uses 6 input nodes, 8 hidden nodes, and 1 output node. The hidden layers use the hyperbolic tangent function, and the output layer uses the sigmoid function. The network expression is:

[0149]

[0150]

[0151] To ensure that the network results simultaneously satisfy both monitoring data and damage evolution patterns, a damage evolution function is defined:

[0152]

[0153] In the formula, Indicates taking The larger of the two values ​​is 0; , and These represent the initial damage stress, initial damage strain, and reference microseismic energy, respectively; let the... The physical feature row vectors corresponding to the group of samples are The difference between damage variables at adjacent time points is used as... Then the parameter vector The least squares solution is:

[0154]

[0155] The parameters identified in this embodiment are:

[0156]

[0157]

[0158] The network training loss function consists of a data fitting term and a physical residual term:

[0159]

[0160] in, The initial learning rate was set to 0.20. The training samples consisted of multi-stage monitoring data of the caisson and supplementary 3D finite element data of the caisson-soil structure output from the monitoring points, totaling 120 sets. Of these, 90 sets were used for model training and 30 sets for independent validation. The Adam optimization algorithm was employed, with an initial learning rate of [missing value]. The maximum number of iterations is 2000; training stops when the verification loss does not decrease for 150 consecutive iterations.

[0161] The model, trained using geometric features from 90 days of operation, yields predicted damage variables. =0.274, and the absolute error between the measured damage variable of 0.277 and the actual damage variable is only 0.003, which verifies the accuracy of the coupled mapping model.

[0162] Step 4: Calculation of stability index

[0163] Based on the output of the coupled mapping model, and combined with the initial overall stiffness and critical instability stiffness of the caisson, a unified expression system of geometry-damage-stiffness degradation is constructed to calculate the current stability index of the caisson.

[0164] Taking the 90D time slot as an example, Figure 3 The geometric features in the table are used as inputs to train the coupled mapping model, resulting in predicted damage variables:

[0165]

[0166] Depend on Figure 4 The equivalent elastic modulus in the table is used to obtain the monitored inversion damage variables:

[0167]

[0168] The absolute error between the two is 0.003; further calculations were performed on the stiffness and stability indices:

[0169]

[0170]

[0171] According to the classification rules The caisson was in a warning state at that time, therefore it remained in a warning state after 90 days of operation; the calculation results for the four typical stages are shown in [link to calculation]. Figure 5 As shown.

[0172] Step 5: Stage Adaptive Weight Calculation

[0173] The evaluation vector for each stage consists of seven parameters: differential settlement, tilt angle, long-side bulging deformation rate, short-side bulging deformation rate, diagonal shear deformation rate, four-corner coordinated deformation coefficient, and damage variable. A 5% single-factor perturbation is applied to each parameter in the four stages of settlement, bottom sealing, backfilling, and operation. The change in stability index is calculated as the sensitivity, and the adaptive weights for each stage are obtained after normalization. The specific steps are as follows:

[0174] Combining geometric features with damage variables to form a stage evaluation vector:

[0175]

[0176] Within each construction phase, the first... Each feature is subjected to a 5% single-factor perturbation, and the change in stability index is calculated by the calibrated model. . No. The stage weights are normalized using the following formula:

[0177]

[0178] The stage-adaptive comprehensive evaluation metric is:

[0179]

[0180] in, The characteristics are normalized according to the design control values; the weights for each stage are shown below. Figure 6 ,Depend on Figure 6 It is evident that the settlement difference and tilt angle are given higher weights during the sinking stage, while the damage variables and the four-corner coordinated deformation coefficient are given higher weights during the operation stage, corresponding to the main sources of risk in different stages.

[0181] Step 6: Time Series Recursive Prediction and Verification

[0182] After 90 days of operation, monitoring results showed that... MPa MPa , ,and Substituting into the damage evolution function:

[0183]

[0184] The damage variables for the next 30 days are predicted using a first-order time recursive formula:

[0185]

[0186] The corresponding stability indicators are:

[0187]

[0188] The prediction results show that if the current damage evolution trend continues, the stability index will be lower than 1.0 after 120 days of operation, entering a state of instability risk. Finite element analysis of the 120-day operation condition yielded a damage variable of 0.303 and a stability index of 0.985. The relative error of the damage prediction was 1.32%, and the relative error of the stability index was 0.51%, indicating that the recursive prediction can provide an early warning before the stability index crosses the risk threshold.

[0189] Method Comparison and Validation

[0190] To verify the effectiveness of the method of this invention, 120 working conditions were constructed, including uniform settlement, uneven settlement, eccentric settlement, local wall stiffness attenuation, changes in bottom sealing constraints, changes in backfill side pressure, and water softening during operation. Figure 2 The placement of the sensors generates equivalent displacement, strain, temperature, and microseismic response; the independent validation set contains 30 working conditions. The method of this invention is compared with a multiple linear regression method using only geometric features and a BP neural network without physical constraint terms; the results are shown in [Figure 1]. Figure 7 As shown.

[0191] Compared to linear regression based on geometric features, the damage prediction MAE of the method in this invention is reduced by 70.8%; compared to a backpropagation neural network without physical constraints, the damage prediction MAE is reduced by 41.8%. Furthermore, the predicted stiffness in the four typical stages shows a consistent trend with the measured stiffness, enabling prediction before the stability index changes from a warning state to a state of instability risk during operation. This demonstrates that by establishing a coupling relationship between geometric shape, internal damage, and stiffness degradation, this invention can improve the accuracy of stability prediction and the timeliness of risk identification for caisson structures.

[0192] The verification data in the above embodiments are numerical working conditions and equivalent sensor outputs established based on specific caisson design parameters, used to illustrate the calculation process and feasibility of the present invention. In practical engineering applications, synchronous monitoring data from the initial stage on-site can be used for recalibration. , , , The damage evolution parameters and network weights remain unchanged, while the rest of the calculation steps remain the same.

[0193] Example 2

[0194] Figure 8 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0195] like Figure 8As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from storage unit 18. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0196] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0197] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the method for predicting the stability of rectangular caisson structures based on geometry-damage coupling mechanisms.

[0198] In some embodiments, the method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism by any other suitable means (e.g., by means of firmware).

[0199] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0200] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0201] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0202] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0203] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0204] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0205] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism, characterized in that, Includes the following steps: (1) Monitoring data acquisition and geometric feature construction: acquire external geometric monitoring data of rectangular caissons during the entire process of sinking, sealing, backfilling and operation, extract multi-dimensional geometric deformation parameters after preprocessing, and construct the geometric stability feature vector of the caisson; (2) Damage and stiffness degradation modeling: collect strain field, temperature field and microseismic energy monitoring data inside the caisson, invert the current equivalent elastic modulus and overall structural stiffness based on the monitoring data, and establish a quantitative relationship model between structural damage variables and stiffness degradation coefficient; (3) Geometric-damage coupling mapping modeling: Using geometric stability feature vectors as input and structural damage variables as output, a nonlinear coupling mapping model with embedded physical constraints of damage evolution is constructed to establish a quantitative correlation between external geometric deformation and internal damage evolution; (4) Stability index calculation: Based on the output of the coupled mapping model, combined with the initial overall stiffness and critical instability stiffness of the caisson, a unified expression system of geometry-damage-stiffness degradation is constructed to calculate the current stability index of the caisson. (5) Stage adaptive weight adjustment: For different construction and operation stages of the caisson, the single-factor sensitivity analysis method is used to determine the adaptive weights of each geometric deformation parameter and damage variable in the corresponding stage, and generate the stage adaptive comprehensive evaluation quantity. (6) Time series prediction and risk classification: Based on the current monitoring data, the damage evolution rate is identified, and the damage variables and stability indicators at future moments are predicted by the time series recursive algorithm. The stability risk level of the caisson structure is classified according to the preset threshold.

2. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (1), the multi-dimensional geometric deformation parameters include: settlement, overall tilt angle, bulging deformation rate of long side panel, bulging deformation rate of short side panel, diagonal shear deformation rate, and four-corner coordination deformation coefficient.

3. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (2), the formula for calculating the structural damage variable is: In the formula, Let E(t) be the damage variable at time t, E0 be the initial elastic modulus, and E(t) be the current equivalent elastic modulus. The formula for calculating the stiffness degradation coefficient is as follows: In the formula, K is the stiffness degradation coefficient at time t, K0 is the initial structural stiffness, K(t) is the current stiffness, and α is the damage influence coefficient. The damage influence coefficient α is obtained by calibration using the least squares method through multiple sets of synchronously monitored damage variables and stiffness data.

4. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (3), the coupling mapping model is constructed using a physical constraint neural network. The network input is the geometric stability feature vector normalized by the design control value, and the network output is the structural damage variable. The training loss function of the physically constrained neural network is composed of a weighted sum of a data fitting term and a physical residual term, wherein the physical residual term is constructed based on the damage evolution differential equation, the expression of which is: In the formula, For damage evolution rate, The initial damage stress, For equivalent stress, For the initial damage strain, For equivalent change, For micro-vibration energy, For reference microseismic energy, , These are the damage evolution parameters to be identified.

5. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, Step (5) specifically includes: All geometric deformation parameters and damage variables are combined into a stage evaluation vector; in each construction or operation stage, a set proportion of single-factor perturbation is applied to each characteristic parameter, and the change in stability index is calculated through the calibrated coupled mapping model, and the change is used as the sensitivity of the corresponding feature. The sensitivity of each feature within the same stage is normalized to obtain the adaptive weights corresponding to each feature in that stage, and the sum of the weights of all features within the same stage is 1. The normalized feature values ​​are summed with their corresponding weights to obtain the adaptive comprehensive evaluation value for this stage.

6. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (6), the time-series recursive algorithm uses a first-order time recursive formula, and the prediction formula for the damage variable at future time steps is: In the formula, For the future Predictive damage variables at time points Let t be the current damage variable at time t. Let be the damage evolution rate at time t. To predict the time step, the stability index for future moments is calculated based on the predicted damage variables, the initial overall stiffness, and the critical buckling stiffness.

7. A method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1 or 6, characterized in that, In step (6), the rule for classifying the stability risk level is as follows: When stability index When the value is ≥1.2, the caisson is considered to be in a safe condition; When 1.0≤ When the value is less than 1.2, the caisson is determined to be in an early warning state; When stability index When the value is less than 1.0, the caisson is determined to be in a state of instability risk. The formula for calculating the stability index is as follows: In the formula, K cr This is the critical instability stiffness of the structure.

8. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (1), the external geometric monitoring data is obtained by combining high-precision displacement monitoring points deployed on the outer wall and top of the caisson and three-dimensional laser scanning equipment; the displacement data is collected at set time intervals, and the three-dimensional point cloud data is collected regularly after each construction stage and during the operation period; The raw monitoring data were preprocessed sequentially through coordinate transformation, outlier removal by three times the standard deviation, and noise reduction by moving average before being used to extract geometric deformation parameters.

9. The method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism according to claim 1, characterized in that, In step (2), the strain field and temperature field are collected by array-type fiber optic sensors at different elevations on the inner side of the caisson, and the mechanical strain is calculated from the total strain by the temperature compensation unit; the micro-vibration energy is continuously collected by the micro-vibration detectors deployed on the sidewall of the caisson and the three-component micro-vibration sensors at the bottom of the caisson.

10. An electronic device, characterized in that, The device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism as described in claim 1.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute the method for predicting the stability of a rectangular caisson structure based on a geometry-damage coupling mechanism as described in claim 1.

Images