Method for predicting upwarp deformation of red-bed soft rock deep excavation foundation in service period

Abstract

The invention discloses a red-layer soft rock deep excavation foundation service period upwarp deformation prediction method, which aims at the problem of continuous upwarp deformation of a deep excavation foundation in a red-layer soft rock region, provides a basis for the upwarp disease treatment of an engineering structure in an operation period by obtaining the maximum value of the upwarp deformation of the foundation in the service period and the time required by deformation stability, and improves the construction efficiency. And meanwhile, reference is provided for disaster prevention design of the deep excavation foundation in the red-layer soft rock area in the future, so that structural damage and economic loss caused by overrun upward arching of the foundation in the service period are avoided. The method can help to achieve continuous monitoring and adjustment of the foundation, effective measures are taken in time according to the pre-judgment result, and the influence of foundation deformation on the upper structure is relieved or prevented. According to the method, reference is provided for design and construction in the construction period, a scientific basis can be provided for treatment of foundation upwarp diseases in the operation period, and safety and reliability of the whole life cycle of a project are ensured.

Description

Technical Field

[0001] The invention relates to the field of red layer soft rock arch deformation prediction, and in particular to a method for predicting the arch deformation of a red layer soft rock deep excavation foundation during its service period. Background Art

[0002] Red-bed soft rock is mainly composed of typical soft rocks such as mudstone, muddy sandstone, sandy mudstone and sandstone. It is widely distributed in southern my country, especially in the southwest. The construction of high-speed railways, highways and high-rise buildings inevitably forms a large number of high-slope road cuttings and deep foundation pits. In the past, it was believed that the rock foundation had good stability and small settlement deformation, and the foundation of the engineering structure would not suffer from deformation during its service life. However, due to the extremely poor engineering properties of red-bed soft rock, low strength, easy weathering, easy disintegration, certain expansibility and significant rheology, especially the unloading effect of excavation, the extremely slow deformation environment conditions under the original geological history time scale have been changed, causing drastic adjustments in the basement stress field; on the other hand, deep excavation will change the original groundwater collection, seepage and drainage conditions, and the physical and mechanical properties of red-bed soft rock will gradually deteriorate during the continuous or cyclic water absorption and dehydration process. In this new stress and water environment, the red-bed soft rock foundation rock mass will cause dynamic mechanical evolution phenomena and processes of deformation and dynamic stress adjustment until it finally reaches a state of equilibrium and deformation stability. Existing studies have shown that the weathering characteristics of red-bed soft rock are extremely sensitive to changes in load and water content. The rheological properties of red-bed soft rock are significantly enhanced under the action of water-mechanical coupling, resulting in significant time-related continuous deformation. In recent years, the red-bed soft rock roadbeds of many high-speed railways and highways in the southwest region, such as the Chengdu-Chongqing High-speed Railway and the Xicheng High-speed Railway, have experienced continuous arching deformation. The continuous arching deformation of the red-bed soft rock building foundation has caused cracks in the basement floor and water seepage.

[0003] The continuous arching deformation of the building foundation will cause cracks in the basement floor, induce basement leakage, water inrush and other diseases, and seriously affect the normal use of the building; the existing high-speed railway ballastless track has an adjustment capacity of only 4.0mm for the roadbed arching. If the arching exceeds the limit, the train must slow down and the track must be broken to adjust the track, causing huge economic losses and social impacts; the long-term continuous arching deformation of the highway roadbed will cause cracks in the road surface, affecting driving safety. However, there are very few research results on the mechanism, law and development trend of the long-term continuous arching deformation of the red layer soft rock roadbed. There is no mature theoretical method at home and abroad that can predict the long-term arching deformation trend, the maximum value of the arching deformation and the stable time of the arching deformation. The prediction of the development trend of the arching deformation during the service life of the foundation is the premise for the design of treatment measures for the superstructure diseases of the red layer soft rock foundation of high-rise buildings, operating high-speed railways, highways, etc. The urban construction of the Chengdu-Chongqing Economic Circle in the western region and a large number of transportation infrastructure construction will face a large number of red layer soft rock foundation projects. Therefore, it is urgent to propose a prediction method for the long-term continuous arching deformation of deep excavation foundations in red-bed soft rock areas to provide a basis for the design of high-rise buildings, high-speed railways, highways and other projects in red-bed soft rock areas. Summary of the invention

[0004] In order to overcome the shortcomings and deficiencies of the existing technology, a method for predicting the arching deformation of deep excavation foundation in red layer soft rock during its service period is proposed.

[0005] A method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period, the method comprising: Step S1, connecting a strain gauge at the confluence node of the red layer soft rock deep excavation foundation to monitor the arch deformation information of the red layer soft rock deep excavation foundation during service, and connecting to the arch deformation remote cloud service computing module through a wireless sensor network; using the deep excavation foundation fiber grating sensor to carry out rock mass deep excavation foundation deformation monitoring, and obtain the deformation data of the deep excavation foundation under different deformation forces; Step S2, determination of the lower limit stress threshold of the rheological flow of the red layer soft rock: collect large blocks of red layer soft rock from the deep excavation foundation base, prepare standard cylindrical samples with a height of 100 mm and a diameter of 50 mm indoors, use the standard samples to carry out indoor lateral water absorption time-dependent expansion deformation tests and graded loading creep tests under water-mechanical coupling conditions, and determine the lower limit stress threshold s of the rheological flow of the red layer soft rock through the graded creep test curve cr ; Step S3, the arch deformation remote cloud service calculation module collects the arch deformation per unit time of the deep excavation foundation, and the arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, and performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation; a high-precision crack meter is used to measure the crack width change of the deep excavation foundation in the process of load deformation in real time, and record the force change of the deep excavation foundation; Step S4, calculation of the time-dependent expansion deformation of the red layer soft rock: using the water absorption expansion deformation-time relationship data obtained from the confined water absorption expansion deformation test of the standard rock sample, the time-dependent expansion deformation parameters K1 and η1 are determined by fitting the time-dependent expansion deformation theoretical model through the least square method, and the maximum water absorption expansion deformation of the rock mass in the time-dependent expansion deformation layer within the thickness range of the roadbed surface layer h1 within the time t1 is calculated; Step S5, determining the base stress state model: using the hydraulic fracturing method, the maximum horizontal stress s at different depths is tested at intervals of 5 m within the range of 10-100 m below the center of the red layer soft rock foundation after excavation. H , the least square method is used to fit the ground stress distribution parameters a, b, c, d within the calculation depth range, and the basement stress state equation is determined; Step S6, the signal transmission module of the arch deformation remote cloud service computing module adopts the communication device module to perform data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation on the arch deformation monitoring platform, perform principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and collect the arch deformation information around the deep excavation foundation of the red layer soft rock, and then upload it to the remote computer control platform after unified data denoising; use a multi-angle high-definition camera to shoot the deep excavation foundation rock layer from multiple angles and viewpoints, and reconstruct the topological structure model of the deep excavation foundation rock layer from the two-dimensional image sequence through the inverse Fourier transform algorithm.

[0006] Furthermore, the method further comprises: Step S7, calculation of rheological deformation of red layer soft rock foundation: the depth below the roadbed surface is h1-h cr The accumulated rheological deformation of the thick rock mass at time t is calculated; the uploaded arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and summarizes the arch deformation information around the deep excavation foundation of the red layer soft rock to the arch deformation remote cloud service calculation module to predict the deformation size and arch deformation state; the DFS topological structure pairing algorithm is used to accurately align the initial deep excavation foundation rock layer topological structure model with the topological structure model after loading; Step S8, calculate the cumulative deformation of the red layer soft rock foundation at any time t during the service period; construct a graph neural network combined with a comparative learning model, and use the acquired crack width changes and generated deformation field data for training and prediction; input the newly acquired deformation field data into the graph neural network prediction model to obtain the corresponding crack width changes, and predict the arch deformation state of the deep excavation foundation according to the predicted crack width changes; Graph Neural Network (GNN) Construction: Define the nodes and edges of the graph. Nodes can be measurement points or key structural points, and edges represent the relationship between these points (such as the degree of mutual influence between stress, strain or cracks). Design the structural layers of the graph neural network, such as GCN (graph convolutional network) or GraphSAGE. Each layer is responsible for learning the characteristics of the node and its neighborhood. Input features: The crack width change and deformation field data are input into the network as node features or edge weights of the graph. Output results: The network output may be the predicted result of the crack width change or the state of foundation deformation.

[0007] Introduction of contrastive learning model: Definition of contrast target: Through contrastive learning, deformation field data can be compared with real crack width change data to learn their potential relationship. Positive and negative sample selection: In the contrastive learning process, positive samples are defined as real deformation field data and its corresponding crack width change, and negative samples are irrelevant data combinations. Loss function: Use the loss function commonly used in contrastive learning (such as contrastive loss or Triplet Loss) for model training, so that positive sample pairs are closer and negative sample pairs are farther apart.

[0008] Model training: Training data: Use preprocessed crack width change data and generated deformation field data for training. Divide the training data into training set, validation set and test set in proportion. Loss function: In addition to the loss function of contrastive learning, you can also add prediction error (such as mean square error) as loss to optimize the crack width prediction accuracy. Optimizer: Use optimization algorithms such as Adam, SGD, etc. to update parameters. Hyperparameter adjustment: Adjust hyperparameters such as learning rate and batch size to optimize training results.

[0009] Prediction and verification: Deformation field data prediction: Input the newly acquired deformation field data into the model to predict the corresponding crack width change. Prediction of foundation deformation state: According to the predicted crack width change, further evaluate the deformation or instability state of the foundation. Model verification and evaluation: Use the verification set to evaluate the model performance, calculate the accuracy, precision, recall and other indicators, and verify whether the model can accurately predict the crack width change and foundation deformation.

[0010] In-depth analysis and optimization: Model optimization: According to the evaluation results, adjust the model structure, hyperparameters or introduce new features to improve the model's prediction performance. Analysis of foundation uplift deformation: According to the predicted crack width changes, combined with physical models or expert knowledge, in-depth analysis of the deformation state of foundation uplift is carried out to determine whether further reinforcement or adjustment of the design is needed.

[0011] Step S9: Use the above formula to draw the camber deformation (S)-time (t) curve of the roadbed during the service period. The camber deformation corresponding to the highest point of the curve is the maximum camber deformation S. sThe moment when the curve tends to be horizontal after the deformation reaches the maximum is the moment when the arch deformation stabilizes t s ; The aligned topological structure model is used for distance calculation, the Euclidean distance formula is used to calculate the deformation variable, and the deformation field drawing software is used to generate the deformation field map of the deep excavation foundation rock layer; the upper arch deformation remote cloud service calculation module is used to predict the deformation size and upper arch deformation state, including deformation residence time analysis prediction, upper arch deformation state level prediction, and maintenance and maintenance measures for the deformed red layer soft rock deep excavation foundation.

[0012] Furthermore, the deformation size and deformation state prediction of the arch deformation remote cloud service calculation module includes the following steps: When the principal component analysis of the upwarping deformation per unit time of the deep excavation foundation changes, the deformation type characteristic analysis prediction program starts; The object arch deformation feature extraction algorithm is used to analyze the collected arch deformation information around the red layer soft rock deep excavation foundation, predict the real-time collected arch deformation-strain relationship, the cumulative impact of the arch deformation, and the peak arch deformation fluctuation of the red layer soft rock deep excavation foundation; predict the changes in the peak-peak value part of the transmission support load of the red layer soft rock deep excavation foundation; predict the partial deformation damage of the red layer soft rock deep excavation foundation and the influence of the damage on the red layer soft rock deep excavation foundation; The arch deformation around the deep excavation foundation of red soft rock, the change of the arch deformation during the service period of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation are used to comprehensively calculate the support and arch deformation dispersion efficiency of the deep excavation foundation of red soft rock. Based on the arch deformation and support of the deep excavation foundation of red soft rock during the service period and the arch deformation dispersion efficiency, the arch deformation of the deep excavation foundation of red soft rock during the service period is analyzed and extracted from the arch deformation during the service period of the rock mass. The overall distribution of the deformation of the deep excavation foundation of red soft rock is calculated using the arch deformation prediction algorithm, and the use environment of the deep excavation foundation of red soft rock is predicted; Based on the above prediction results, the status of deep excavation foundation in red layer soft rock is graded and evaluated, and different foundation maintenance measures are given.

[0013] Furthermore, the specific steps of the object arch deformation feature extraction algorithm are as follows: Establish a database of the status of deep excavation foundations in red layer soft rock under different weather conditions, including a waveform database of normal conditions of deep excavation foundations in red layer soft rock with different levels of support loads, a historical waveform database of tested deep excavation foundations in red layer soft rock, and waveform databases of various defects of deep excavation foundations in red layer soft rock; Preprocessing the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; Feature extraction and selection are performed on the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; The characteristic support load curve of deep excavation foundation in red layer soft rock is generated, and the characteristic matching and arch deformation are compared and analyzed. By quantifying the difference between the measured waveform and the normal waveform and historical waveform, the type and state of deformation are predicted, and the deformation defects and causes are predicted by matching with the defect waveform characteristics.

[0014] Furthermore, the specific steps of using the arch deformation prediction algorithm to calculate the rock mass crack distribution of the red layer soft rock deep excavation foundation and predicting the use environment of the red layer soft rock deep excavation foundation are as follows: Monitor the maximum deformation of deep excavation foundation of red soft rock and the crack depth parameters of rock mass; The support and dispersion efficiency of the red layer soft rock deep excavation foundation are comprehensively calculated by using the arch deformation information around the deep excavation foundation of the red layer soft rock, the changes in the arch deformation during the service life of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation; Based on the arching deformation and support of deep excavation foundation in red layer during its service period and the dispersion efficiency of the arching deformation, the arching deformation of deep excavation foundation in red layer during its service period is analyzed and extracted from the arching deformation of rock mass during its service period, and the overall distribution of deep excavation foundation deformation in red layer during its service period is calculated using the arching deformation prediction algorithm. The overall distribution of deformation of the deep excavation foundation in red layer soft rock is used to predict the use environment of the deep excavation foundation and the degree of damage to the deep excavation foundation in red layer soft rock.

[0015] Furthermore, the arch deformation information of the red layer soft rock deep excavation foundation during service period monitored by the strain gauge is converted into principal component analysis arch deformation by a mechanical arch deformation converter, and then connected to the arch deformation remote cloud service calculation module through a wireless sensor network.

[0016] Furthermore, in step S2, the maximum water absorption expansion deformation expression is: (1) Among them, S1(t) represents the time-related expansion deformation, h1 represents the thickness of the rock layer below the roadbed surface that produces time-dependent expansion deformation, and the corresponding thickness of the atmospheric influence layer of the expansion rock is taken according to the meteorological conditions in the project area, with a value range of 1.0-2.0m; t represents the duration, K1 represents the limiting expansion strain parameter; η1 represents the viscosity coefficient of water absorption expansion, and e represents the natural constant.

[0017] Furthermore, in step S3, the substrate stress state equation is expressed as: (2) Among them, s H represents the maximum horizontal principal stress; s Vrepresents vertical stress; z represents the distance from the top of the roadbed (m); λ represents the lateral pressure coefficient of the rock mass, which can be taken as 0.5 in areas without obvious tectonic stress influence; γ represents the average weight of the red layer soft rock (kN / m 3 ), determined according to the results of the on-site rock density test; a, b, c, d are the characteristic parameters of the ground stress distribution to be determined.

[0018] Furthermore, the determined base stress state equation (2) is used to solve the following equation (3) to determine the calculated depth z and obtain the critical calculated depth h of the water-mechanical coupling rheological layer rock mass: cr =z, (s H -s V -P v )-s cr =0 (3) Among them, P v It refers to the vertical external load acting on the foundation. For deep foundation pits of high-rise buildings, it is the pressure transmitted from the superstructure to the bottom of the foundation. For deep cutting roadbeds such as high-speed railways and highways, it is the pressure transmitted to the top of the foundation from the deadweight of the roadbed body and train loads.

[0019] Furthermore, in step S4, the expression of the accumulated rheological deformation at time t is: (4) Among them, G 2e is the equivalent viscoelastic shear modulus of the rheological layer rock mass, η 1e and η 2e is the equivalent viscosity coefficient of the rheological layer rock mass, which reflects the constant creep rate and attenuated creep rate of the rock, respectively. The equivalent rheological parameters of the three rheological layers are obtained by fitting the Burgers model using the least squares method based on the indoor hydraulic-mechanical coupling graded loading creep test data of the rock sample. v and σ H are the vertical stress and horizontal stress values ​​of the base related to the depth z, respectively, and are determined by formula (2).

[0020] Furthermore, in step S5, the cumulative deformation at any time t during the service period is: (5). Beneficial Effects

[0021] The present invention proposes a method for predicting the arch deformation of deep excavation foundation in red layer soft rock during the service period. The method aims at the problem of continuous arch deformation of deep excavation foundation in red layer soft rock area. By obtaining the maximum value of the arch deformation of the foundation during the service period and the time required for deformation stabilization, it provides a basis for the treatment of arch disease of engineering structure during the operation period. It also provides a reference for the disaster prevention design of deep excavation foundation in red layer soft rock area in the future, so as to avoid structural damage and economic losses caused by excessive arching of foundation during the service period. The method can help to realize the continuous monitoring and adjustment of the foundation, and take effective measures in time through the prediction results to reduce or prevent the impact of foundation deformation on the superstructure. The method not only provides a reference for the design and construction during the construction period, but also provides a scientific basis for the treatment of arch disease of the foundation during the operation period, ensuring the safety and reliability of the whole life cycle of the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic flow chart of the method of the present invention; Figure 2 This is a model diagram of a deep excavation foundation of red layer soft rock of the present invention; Figure 3 This is a long-term deformation prediction model diagram of a red layer soft rock deep excavation foundation of the present invention; Figure 4 It is a roadbed deformation curve diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present invention will be further described in detail and completely below in conjunction with the drawings in the embodiments of the present invention. It is obvious that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0024] The embodiment of the present invention discloses a method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during its service period. The method obtains the maximum value of the foundation arching deformation and the time required for deformation stabilization during the service period, providing a basis for the treatment of arching diseases of engineering structures during the operation period. It also provides a reference for the disaster prevention design of deep excavation foundations in red layer soft rock areas in the future, so as to avoid structural damage and economic losses caused by excessive arching of the foundation during the service period.

[0025] like Figure 1 As shown, a method for predicting the uplift deformation of a deep excavation foundation in red layer soft rock during service period comprises: Step S1, connecting a strain gauge at the confluence node of the red layer soft rock deep excavation foundation to monitor the arch deformation information of the red layer soft rock deep excavation foundation during service, and connecting to the arch deformation remote cloud service computing module through a wireless sensor network; using the deep excavation foundation fiber grating sensor to carry out rock mass deep excavation foundation deformation monitoring, and obtain the deformation data of the deep excavation foundation under different deformation forces; Step S2, determination of the lower limit stress threshold of the rheological flow of the red layer soft rock: collect large blocks of red layer soft rock from the deep excavation foundation base, prepare standard cylindrical samples with a height of 100 mm and a diameter of 50 mm indoors, use the standard samples to carry out indoor lateral water absorption time-dependent expansion deformation tests and graded loading creep tests under water-mechanical coupling conditions, and determine the lower limit stress threshold s of the rheological flow of the red layer soft rock through the graded creep test curve cr ; Step S3, the arch deformation remote cloud service calculation module collects the arch deformation per unit time of the deep excavation foundation, and the arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, and performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation; a high-precision crack meter is used to measure the crack width change of the deep excavation foundation in the process of load deformation in real time, and record the force change of the deep excavation foundation; Step S4, calculation of the time-dependent expansion deformation of the red layer soft rock: using the water absorption expansion deformation-time relationship data obtained from the confined water absorption expansion deformation test of the standard rock sample, the time-dependent expansion deformation parameters K1 and η1 are determined by fitting the time-dependent expansion deformation theoretical model through the least square method, and the maximum water absorption expansion deformation of the rock mass in the time-dependent expansion deformation layer within the thickness range of the roadbed surface layer h1 within the time t1 is calculated; Step S5, determining the base stress state model: using the hydraulic fracturing method, the maximum horizontal stress s at different depths is tested at intervals of 5 m within the range of 10-100 m below the center of the red layer soft rock foundation after excavation. H , the following formula is used to fit the least square method to obtain the ground stress distribution parameters a, b, c, d within the calculation depth range, and determine the basement stress state equation; Step S6, the signal transmission module of the arch deformation remote cloud service computing module adopts the communication device module to perform data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation on the arch deformation monitoring platform, perform principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and collect the arch deformation information around the deep excavation foundation of the red layer soft rock, and then upload it to the remote computer control platform after unified data denoising; use a multi-angle high-definition camera to shoot the deep excavation foundation rock layer from multiple angles and viewpoints, and reconstruct the topological structure model of the deep excavation foundation rock layer from the two-dimensional image sequence through the inverse Fourier transform algorithm.

[0026] Furthermore, the method further comprises: Step S7, calculation of rheological deformation of red layer soft rock foundation: the depth below the roadbed surface is h1-hcr The accumulated rheological deformation of the thick rock mass at time t is calculated; the uploaded arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and summarizes the arch deformation information around the deep excavation foundation of the red layer soft rock to the arch deformation remote cloud service calculation module to predict the deformation size and arch deformation state; the DFS topological structure pairing algorithm is used to accurately align the initial deep excavation foundation rock layer topological structure model with the topological structure model after loading; Step S8, calculate the cumulative deformation of the red layer soft rock foundation at any time t during the service period; construct a graph neural network combined with a comparative learning model, and use the acquired crack width changes and generated deformation field data for training and prediction; input the newly acquired deformation field data into the graph neural network prediction model to obtain the corresponding crack width changes, and predict the arch deformation state of the deep excavation foundation according to the predicted crack width changes; Step S9: Use the above formula to draw the camber deformation (S)-time (t) curve of the roadbed during the service period. The camber deformation corresponding to the highest point of the curve is the maximum camber deformation S. s The moment when the curve tends to be horizontal after the deformation reaches the maximum is the moment when the arch deformation stabilizes t s ; The aligned topological structure model is used for distance calculation, the Euclidean distance formula is used to calculate the deformation variable, and the deformation field drawing software is used to generate the deformation field map of the deep excavation foundation rock layer; the upper arch deformation remote cloud service calculation module is used to predict the deformation size and upper arch deformation state, including deformation residence time analysis prediction, upper arch deformation state level prediction, and maintenance and maintenance measures for the deformed red layer soft rock deep excavation foundation.

[0027] Furthermore, the deformation size and deformation state prediction of the arch deformation remote cloud service calculation module includes the following steps: When the principal component analysis of the upwarping deformation per unit time of the deep excavation foundation changes, the deformation type characteristic analysis prediction program starts; The object arch deformation feature extraction algorithm is used to analyze the collected arch deformation information around the red layer soft rock deep excavation foundation, predict the real-time collected arch deformation-strain relationship, the cumulative impact of the arch deformation, and the peak arch deformation fluctuation of the red layer soft rock deep excavation foundation; predict the changes in the peak-peak value part of the transmission support load of the red layer soft rock deep excavation foundation; predict the partial deformation damage of the red layer soft rock deep excavation foundation and the influence of the damage on the red layer soft rock deep excavation foundation; The arch deformation around the deep excavation foundation of red soft rock, the change of the arch deformation during the service period of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation are used to comprehensively calculate the support and arch deformation dispersion efficiency of the deep excavation foundation of red soft rock. Based on the arch deformation and support of the deep excavation foundation of red soft rock during the service period and the arch deformation dispersion efficiency, the arch deformation of the deep excavation foundation of red soft rock during the service period is analyzed and extracted from the arch deformation during the service period of the rock mass. The overall distribution of the deformation of the deep excavation foundation of red soft rock is calculated using the arch deformation prediction algorithm, and the use environment of the deep excavation foundation of red soft rock is predicted; Based on the above prediction results, the status of deep excavation foundation in red layer soft rock is graded and evaluated, and different foundation maintenance measures are given.

[0028] Furthermore, the specific steps of the object arch deformation feature extraction algorithm are as follows: Establish a database of the status of deep excavation foundations in red layer soft rock under different weather conditions, including a waveform database of normal conditions of deep excavation foundations in red layer soft rock with different levels of support loads, a historical waveform database of tested deep excavation foundations in red layer soft rock, and waveform databases of various defects of deep excavation foundations in red layer soft rock; Preprocessing the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; Feature extraction and selection are performed on the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; The characteristic support load curve of deep excavation foundation in red layer soft rock is generated, and the characteristic matching and arch deformation are compared and analyzed. By quantifying the difference between the measured waveform and the normal waveform and historical waveform, the type and state of deformation are predicted, and the deformation defects and causes are predicted by matching with the defect waveform characteristics.

[0029] Furthermore, the specific steps of using the arch deformation prediction algorithm to calculate the rock mass crack distribution of the red layer soft rock deep excavation foundation and predicting the use environment of the red layer soft rock deep excavation foundation are as follows: Monitor the maximum deformation of the deep excavation foundation of red soft rock and the crack depth parameters of the rock mass; The support and dispersion efficiency of the red layer soft rock deep excavation foundation are comprehensively calculated by using the arch deformation information around the deep excavation foundation of the red layer soft rock, the changes in the arch deformation during the service life of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation; Based on the arching deformation and support of deep excavation foundation in red layer during its service period and the dispersion efficiency of the arching deformation, the arching deformation of deep excavation foundation in red layer during its service period is analyzed and extracted from the arching deformation of rock mass during its service period, and the overall distribution of deep excavation foundation deformation in red layer during its service period is calculated using the arching deformation prediction algorithm. The overall distribution of deformation of the deep excavation foundation in red layer soft rock is used to predict the use environment of the deep excavation foundation and the degree of damage to the deep excavation foundation in red layer soft rock.

[0030] Furthermore, the arch deformation information of the red layer soft rock deep excavation foundation during service period monitored by the strain gauge is converted into principal component analysis arch deformation by a mechanical arch deformation converter, and then connected to the arch deformation remote cloud service calculation module through a wireless sensor network.

[0031] like Figure 2 As shown in the figure, the structure of the deep excavation foundation and the related deformation are described as follows: Top: The top of the diagram shows the "subgrade," or slope of the ground, which indicates that above the foundation is a roadbed or similar structure.

[0032] h1: It indicates the height from the roadbed surface to the time-dependent expansion deformation layer. This area is the upper layer of the foundation and may be affected by the time-dependent expansion effect.

[0033] Time-dependent expansion deformation layer: This area is located below h1, indicating that this is the area where foundation deformation occurs under the action of time, which may be caused by material properties or environmental factors.

[0034] h2: The height from the aging expansion deformation layer to the rheological deformation layer at the bottom.

[0035] Rheological deformation layer: Located deeper in the foundation, it is marked as the area where "rheological deformation" occurs. Rheological deformation is the gradual deformation of the material due to long-term stress.

[0036] Base (h cr ):h cr It indicates the depth from the ground surface to the base, and identifies the excavation depth or bearing depth of the entire foundation.

[0037] Left and right boundaries: The left and right sides of the figure represent fixed boundaries, which are the constraints of the structure.

[0038] Figure 2 The two different deformation layers of time-dependent expansion deformation and rheological deformation in deep excavation foundation are illustrated, showing how the deformation of the foundation changes with depth.

[0039] like Figure 3 As shown, a cross-sectional diagram related to foundation deformation is shown, combining rainfall effects and stress distribution to explain the deformation and stress changes of the roadbed under different conditions. The following is a description of each part of the figure: Middle part (foundation section): Road foundation volume: The upper part of the figure shows the "road foundation volume", which represents the surface part of the foundation.

[0040] C1: Time-dependent expansion deformation layer: The first layer under the roadbed is the time-dependent expansion deformation layer, which means that this layer will produce expansion deformation over time, which may be caused by soil characteristics or environmental changes.

[0041] C2: Water-mechanical coupled rheological layer: Located below C1 is the water-mechanical coupled rheological layer, which may produce rheological deformation due to the action of moisture and long-term stress.

[0042] h1 and h2: h1 is the thickness of the time-dependent expansion deformation layer, h2 is the thickness of the water-mechanical coupling rheological deformation layer, and h0 is the total depth from the surface to the basement.

[0043] Left part (deformation curve): s, s1, s2: This is a schematic curve of foundation deformation, the horizontal axis represents the deformation (s), and the vertical axis represents the depth (z). s1 and s2 represent the deformation at different depths.

[0044] Right part (stress distribution): Stress distribution curve: shows the distribution of stress in the foundation, the horizontal axis represents stress (σ) and the vertical axis represents depth (z).

[0045] Stress Concentration Area: Stress concentration occurs at shallower locations and may be caused by foundation expansion or external loads.

[0046] Stress increase zone: As the depth increases, the stress gradually increases until it reaches a certain maximum value (Δσmax).

[0047] Stress difference formula: Stress difference Δσ = σ H (t,h) - σ v (h) describes the difference between horizontal and vertical stresses in the foundation, which shows the changing relationship between horizontal and vertical stresses at different depths.

[0048] Effect of rainfall: The figure shows a symbol of a rainy cloud, indicating that rainfall is an important factor affecting the deformation of the foundation, especially the deformation behavior of the water-mechanical coupling deformation layer.

[0049] Figure 3 Through the cross-section diagram and stress distribution curve, the stress and deformation distribution of the roadbed and the different deformation layers underneath it under rainfall conditions are described, reflecting the response behavior of the foundation under external loads and environmental changes.

[0050] Preferably, in step S2, the maximum water absorption expansion deformation expression is: (1) Among them, S1(t) represents the time-related expansion deformation, h1 represents the thickness of the rock layer below the roadbed surface that produces time-dependent expansion deformation, and the corresponding thickness of the atmospheric influence layer of the expansion rock is taken according to the meteorological conditions in the project area, with a value range of 1.0-2.0m; t represents the duration, K1 represents the limiting expansion strain parameter; η1 represents the viscosity coefficient of water absorption expansion, and e represents the natural constant.

[0051] Preferably, in step S3, the substrate stress state equation is expressed as: (2) Among them, s H represents the maximum horizontal principal stress; s V represents vertical stress; z represents the distance from the top of the roadbed (m); λ represents the lateral pressure coefficient of the rock mass, which can be taken as 0.5 in areas without obvious tectonic stress influence; γ represents the average weight of the red layer soft rock (kN / m 3 ), determined according to the results of the on-site rock density test; a, b, c, d are the characteristic parameters of the ground stress distribution to be determined.

[0052] Preferably, the determined base stress state equation (2) is used to solve the following equation (3) to determine the calculated depth z, and the critical calculated depth h of the water-mechanical coupling rheological layer rock mass is obtained: cr =z, (s H -s V -P v )-s cr =0 (3) Among them, P v It refers to the vertical external load acting on the foundation. For deep foundation pits of high-rise buildings, it is the pressure transmitted from the superstructure to the bottom of the foundation. For deep cutting roadbeds such as high-speed railways and highways, it is the pressure transmitted to the top of the foundation from the deadweight of the roadbed body and train loads.

[0053] Preferably, in step S4, the cumulative rheological deformation expression at time t is: (4) Among them, G 2e is the equivalent viscoelastic shear modulus of the rheological layer rock mass, η 1e and η 2e is the equivalent viscosity coefficient of the rheological layer rock mass, which reflects the constant creep rate and attenuated creep rate of the rock, respectively. The equivalent rheological parameters of the three rheological layers are obtained by fitting the Burgers model using the least squares method based on the indoor hydraulic-mechanical coupling graded loading creep test data of the rock sample. v and σ H are the vertical stress and horizontal stress values ​​of the base related to the depth z, respectively, and are determined by formula (2).

[0054] The prediction results of arch deformation are obtained by fitting the Burgers model using the least squares method, which can more accurately describe the viscoelastic-plastic deformation behavior of the red-bed soft rock foundation. The Burgers model combines viscous and elastic elements, which can better simulate the progressive deformation characteristics of soft rock under long-term loads, and the least squares rule is used to reduce the error in the fitting process, thereby improving the accuracy of the model. This method can effectively predict the arch deformation of the foundation during its service life, provide more reliable maximum deformation and deformation stabilization time, and provide a scientific basis for engineering design and disease treatment.

[0055] Preferably, in step S5, the cumulative deformation at any time t during the service period is: (5).

[0056] Preferably, in a high-speed railway red layer soft rock deep excavation cutting foundation, the expansion and rheological parameters of the base rock mass are determined by indoor tests and on-site ground stress tests according to the above steps 1-2 as shown in the following table:

[0057] Taking the expansion influence layer depth h1 = 2000mm, the average rock mass density γ = 25kN / m3, and the rheological critical stress threshold scr = 100kPa according to the rock sample water-mechanical coupling graded loading creep test results, the critical calculation depth hcr = 53m is obtained by combining equations (2) and (3).

[0058] The above parameters determine formula (1) as:

[0059] Formula (2) is:

[0060] Formula (4) is:

[0061] The above formula can be used to draw the arch deformation (S)-time (t) curve: like Figure 4 As shown in the figure above, it can be determined that the maximum predicted maximum arch deformation of the red layer soft rock roadbed during the operation period is 69.4mm, and the time to reach stable deformation is about 10 years.

[0062] Although the embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period, characterized in that: The method includes: Step S1, connecting a strain gauge at the confluence node of the red layer soft rock deep excavation foundation to monitor the arch deformation information of the red layer soft rock deep excavation foundation during service, and connecting to the arch deformation remote cloud service computing module through a wireless sensor network; using the deep excavation foundation fiber grating sensor to carry out rock mass deep excavation foundation deformation monitoring, and obtain the deformation data of the deep excavation foundation under different deformation forces; Step S2, determination of the lower limit stress threshold of the rheological flow of the red layer soft rock: collect large blocks of red layer soft rock from the deep excavation foundation base, prepare standard cylindrical samples with a height of 100 mm and a diameter of 50 mm indoors, use the standard samples to carry out indoor lateral water absorption time-dependent expansion deformation tests and graded loading creep tests under water-mechanical coupling conditions, and determine the lower limit stress threshold s of the rheological flow of the red layer soft rock through the graded creep test curve cr ; Step S3, the arch deformation remote cloud service calculation module collects the arch deformation per unit time of the deep excavation foundation, and the arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, and performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation; a high-precision crack meter is used to measure the crack width change of the deep excavation foundation in the process of load deformation in real time, and record the force change of the deep excavation foundation; Step S4, calculation of the time-dependent expansion deformation of the red layer soft rock: using the water absorption expansion deformation-time relationship data obtained from the confined water absorption expansion deformation test of the standard rock sample, the time-dependent expansion deformation parameters K1 and η1 are determined by fitting the time-dependent expansion deformation theoretical model through the least square method, and the maximum water absorption expansion deformation of the rock mass in the time-dependent expansion deformation layer within the thickness range of the roadbed surface layer h1 within the time t1 is calculated; Step S5, determining the base stress state model: using the hydraulic fracturing method, the maximum horizontal stress s at different depths is tested at intervals of 5 m within the range of 10-100 m below the center of the red layer soft rock foundation after excavation. H , the least square method is used to fit the ground stress distribution parameters a, b, c, d within the calculation depth range, and the basement stress state equation is determined; Step S6, the signal transmission module of the arch deformation remote cloud service computing module adopts the communication device module to perform data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation on the arch deformation monitoring platform, perform principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and collect the arch deformation information around the deep excavation foundation of the red layer soft rock, and then upload it to the remote computer control platform after unified data denoising; use a multi-angle high-definition camera to shoot the deep excavation foundation rock layer from multiple angles and viewpoints, and reconstruct the topological structure model of the deep excavation foundation rock layer from the two-dimensional image sequence through the inverse Fourier transform algorithm.

2. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period as claimed in claim 1, characterized in that: The method further includes: Step S7, calculation of rheological deformation of red layer soft rock foundation: the depth below the roadbed surface is h1-h cr The accumulated rheological deformation of the thick rock mass at time t is calculated; the uploaded arch deformation monitoring platform performs data desensitization and deduplication processing on the arch deformation per unit time of the deep excavation foundation, performs principal component analysis processing on the arch deformation per unit time of the deep excavation foundation, and summarizes the arch deformation information around the deep excavation foundation of the red layer soft rock to the arch deformation remote cloud service calculation module to predict the deformation size and arch deformation state; the DFS topological structure pairing algorithm is used to accurately align the initial deep excavation foundation rock layer topological structure model with the topological structure model after loading; Step S8, calculate the cumulative deformation of the red layer soft rock foundation at any time t during the service period; construct a graph neural network combined with a comparative learning model, and use the acquired crack width changes and generated deformation field data for training and prediction; input the newly acquired deformation field data into the graph neural network prediction model to obtain the corresponding crack width changes, and predict the arch deformation state of the deep excavation foundation according to the predicted crack width changes; Step S9: Use the above formula to draw the camber deformation (S)-time (t) curve of the roadbed during the service period. The camber deformation corresponding to the highest point of the curve is the maximum camber deformation S. s The moment when the curve tends to be horizontal after the deformation reaches the maximum is the moment when the arch deformation stabilizes t s ; The aligned topological structure model is used for distance calculation, the Euclidean distance formula is used to calculate the deformation variable, and the deformation field drawing software is used to generate the deformation field map of the deep excavation foundation rock layer; the upper arch deformation remote cloud service calculation module is used to predict the deformation size and upper arch deformation state, including deformation residence time analysis prediction, upper arch deformation state level prediction, and maintenance and maintenance measures for the deformed red layer soft rock deep excavation foundation.

3. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period according to claim 1 is characterized in that: The arch deformation remote cloud service calculation module predicts the deformation size and the arch deformation state, including the following steps: When the principal component analysis of the upwarping deformation per unit time of the deep excavation foundation changes, the deformation type characteristic analysis prediction program starts; The object arch deformation feature extraction algorithm is used to analyze the collected arch deformation information around the red layer soft rock deep excavation foundation, predict the real-time collected arch deformation-strain relationship, the cumulative impact of the arch deformation, and the peak arch deformation fluctuation of the red layer soft rock deep excavation foundation; predict the changes in the peak-peak value part of the transmission support load of the red layer soft rock deep excavation foundation; predict the partial deformation damage of the red layer soft rock deep excavation foundation and the influence of the damage on the red layer soft rock deep excavation foundation; The arch deformation around the deep excavation foundation of red soft rock, the change of the arch deformation during the service period of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation are used to comprehensively calculate the support and arch deformation dispersion efficiency of the deep excavation foundation of red soft rock. Based on the arch deformation and support of the deep excavation foundation of red soft rock during the service period and the arch deformation dispersion efficiency, the arch deformation of the deep excavation foundation of red soft rock during the service period is analyzed and extracted from the arch deformation during the service period of the rock mass. The overall distribution of the deformation of the deep excavation foundation of red soft rock is calculated using the arch deformation prediction algorithm, and the use environment of the deep excavation foundation of red soft rock is predicted; Based on the above prediction results, the status of deep excavation foundation in red layer soft rock is graded and evaluated, and different foundation maintenance measures are given.

4. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period according to claim 3 is characterized in that: The specific steps of the object arch deformation feature extraction algorithm are as follows: Establish a database of the status of deep excavation foundations in red layer soft rock under different weather conditions, including a waveform database of normal conditions of deep excavation foundations in red layer soft rock with different levels of support loads, a historical waveform database of tested deep excavation foundations in red layer soft rock, and waveform databases of various defects of deep excavation foundations in red layer soft rock; Preprocessing the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; Feature extraction and selection are performed on the collected arch deformation waveform information of the deep excavation foundation of red layer soft rock during service period; The characteristic support load curve of deep excavation foundation in red layer soft rock is generated, and the characteristic matching and arch deformation are compared and analyzed. By quantifying the difference between the measured waveform and the normal waveform and historical waveform, the type and state of deformation are predicted, and the deformation defects and causes are predicted by matching with the defect waveform characteristics.

5. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period according to claim 3 is characterized in that: The specific steps of using the arch deformation prediction algorithm to calculate the rock mass crack distribution of the red layer soft rock deep excavation foundation and predicting the use environment of the red layer soft rock deep excavation foundation are as follows: Monitor the maximum deformation of the deep excavation foundation of red soft rock and the crack depth parameters of the rock mass; The support and dispersion efficiency of the red layer soft rock deep excavation foundation are comprehensively calculated by using the arch deformation information around the deep excavation foundation of the red layer soft rock, the changes in the arch deformation during the service life of the rock mass, and the principal component analysis of the arch deformation per unit time of the deep excavation foundation; Based on the arching deformation and support of deep excavation foundation in red layer during its service period and the dispersion efficiency of the arching deformation, the arching deformation of deep excavation foundation in red layer during its service period is analyzed and extracted from the arching deformation of rock mass during its service period, and the overall distribution of deep excavation foundation deformation in red layer during its service period is calculated using the arching deformation prediction algorithm. The overall distribution of deformation of the deep excavation foundation in red layer soft rock is used to predict the use environment of the deep excavation foundation and the degree of damage to the deep excavation foundation in red layer soft rock.

6. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period according to claim 1 is characterized in that: The arch deformation information of the red layer soft rock deep excavation foundation during service period monitored by the strain gauge is converted into principal component analysis arch deformation by a mechanical arch deformation converter, and then connected to the arch deformation remote cloud service calculation module through a wireless sensor network.

7. The method for predicting the arching deformation of a deep excavation foundation in red layer soft rock during service period according to claim 1 is characterized in that: In step S5, the substrate stress state equation is expressed as: in, represents the maximum horizontal principal stress; represents vertical stress; z represents the distance from the top of the roadbed (m); λ represents the lateral pressure coefficient of the rock mass, which can be taken as 0.5 in areas without obvious tectonic stress influence; γ represents the average weight of the red layer soft rock (kN / m 3 ), determined according to the results of the on-site rock density test; a, b, c, d are the characteristic parameters of the ground stress distribution to be determined.

Images