Rock soil stress monitoring data analysis method and system

Abstract

The invention relates to the technical field of rock-soil stress monitoring, in particular to a rock-soil stress monitoring data analysis method and system. The method comprises the following steps: acquiring geological data and construction design data of a river bridge construction area, and analyzing piling immersed tube distribution difference to obtain piling immersed tube distribution difference data; performing extrusion lateral stress deflection simulation according to the piling immersed tube distribution difference data, further estimating stress transfer loss to obtain effective stress transfer data, and performing stress impact structure instability analysis between adjacent rocks and soil; and constructing a rock-soil stress risk monitoring model based on a random forest algorithm, and sending the model to a terminal for rock-soil stress monitoring data analysis. The rock-soil stress monitoring technology is optimized, so that the rock-soil stress monitoring technology is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of rock and soil stress monitoring, and in particular to a rock and soil stress monitoring data analysis method and system. Background Art

[0002] Geotechnical engineering occupies an important position in modern infrastructure construction, especially in the construction process of large-scale projects such as bridges, tunnels, and dams. The stress state of geotechnical engineering directly affects the safety and stability of the project. With the continuous expansion of the scale of the project and the increasing complexity of the construction environment, the deformation, stress distribution and mechanical properties of the geotechnical body have become more difficult to predict. Especially under special geological conditions such as river bridges, the construction often involves complex engineering measures such as piling and sinking pipes. These operations will cause stress changes, settlement and displacement of the surrounding rock and soil, and even cause structural instability or damage of adjacent rock and soil layers, seriously affecting the safety of the project. Therefore, timely and accurate monitoring of geotechnical stress changes and effective analysis have become an important part of ensuring project safety. However, a traditional geotechnical stress monitoring data analysis method has the problem of low accuracy in identifying structural instability caused by stress impact, resulting in large errors in geotechnical stress monitoring. Summary of the invention

[0003] Based on this, it is necessary to provide a geotechnical stress monitoring data analysis method and system to solve at least one of the above technical problems.

[0004] To achieve the above object, a geotechnical stress monitoring data analysis method is provided, the method comprising the following steps: Step S1: obtaining geological data of the bridge construction area on the river and design data of the river bridge construction; summarizing the differences in the distribution of pile driving and immersed pipes on the river bridge construction design data to obtain data on the differences in the distribution of pile driving and immersed pipes; Step S2: simulate the lateral stress deflection of pile driving extrusion according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transfer loss of the lateral stress deflection data of extrusion to obtain the effective stress transfer data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transfer data, and obtain the stress impact structural instability data between adjacent rocks and soils; Step S3: construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm to obtain a geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0005] Preferably, step S1 comprises the following steps: Step S11: Acquire geological data of the bridge construction area on the river and design data of the river bridge construction; Step S12: marking the river bridge construction design data with a construction time series to obtain construction design time marked data; Step S13: performing piling sequence selection analysis on the construction design time-stamped data to obtain construction piling sequence selection data; Step S14: Summarize the differences in the distribution of pile driving and pipe immersion according to the construction piling sequence selection data and obtain the piling and pipe immersion distribution difference data.

[0006] Preferably, step S2 comprises the following steps: Step S21: performing vertical geological and geotechnical structure analysis of different piling and immersion pipe areas on the geological data of the bridge construction area based on the piling and immersion pipe distribution difference data to obtain vertical geological and geotechnical structure data of different piling and immersion pipe areas; Step S22: simulating the lateral stress deflection caused by pile driving according to the pile driving and pipe sinking distribution difference data on the vertical geological and geotechnical structure data to obtain the lateral stress deflection data caused by pile driving; Step S23: identifying the stress radial influence orientation on the extrusion lateral stress deflection data to obtain the stress radial influence orientation; Step S24: estimating stress transfer loss based on the stress radial influence orientation to obtain effective stress transfer data; Step S25: Based on the effective stress transfer data, the vertical geological rock and soil structure data is subjected to stress impact structural instability analysis between adjacent rocks and soils to obtain stress impact structural instability data between adjacent rocks and soils.

[0007] Preferably, step S22 includes the following steps: Step S221: performing force / frequency analysis on the pile driving and immersed pipe distribution difference data to obtain the pile driving force / frequency data; Step S222: Perform vertical layered density analysis on the vertical geological rock and soil structure data to obtain the vertical layered density of the rock and soil; Step S223: Identify the lateral contact section rotation angle of the vertical layered density of the rock and soil according to the force / frequency data of the pile driving, and obtain the contact section rotation angle; Step S224: evaluating the fluctuation of the internal friction strength of the rock and soil between different layer depths of the vertical layered density of the rock and soil based on the pile driving force / frequency data and the contact section rotation angle, and obtaining the fluctuation data of the internal friction strength of the rock and soil; Step S225: performing strength boundary fluctuation integration on the rock and soil internal friction strength fluctuation data to obtain the internal friction boundary fluctuation integral value; Step S226: solving the boundary-constrained internal friction interval for the rock and soil internal friction intensity fluctuation data based on the variational method and the internal friction boundary fluctuation integral value to obtain the boundary-constrained internal friction interval; Step S227: performing a simulation of the lateral stress deflection caused by pile driving extrusion according to the friction force interval within the boundary constraint to obtain the extrusion lateral stress deflection data.

[0008] Preferably, step S25 comprises the following steps: Step S251: performing geotechnical classification on the vertical geological and geotechnical structure data to obtain vertical geological and geotechnical classification types; Step S252: Based on the vertical geological and rock-soil classification type, the vertical geological and rock-soil structure data is subjected to geological fold / fault profile morphology recognition between adjacent rocks and soils to obtain geological fold / fault profile morphology data; Step S253: performing stress progressive superposition frequency evaluation on the effective stress transfer data to obtain stress progressive superposition data; performing fold deformation inclination angle calculation on the geological fold / fault profile morphology data to obtain the fold deformation inclination angle; Step S254: calculating the average difference of fold surface slopes at different depths for the geological fold / fault profile morphology data according to the fold deformation inclination angle, and obtaining the average difference of fold surface slopes; Step S255: Based on the stress progressive superposition data, a stress impact structural instability simulation is performed on the fold deformation inclination angle and the fold surface slope difference to obtain stress impact fold structural instability data; Step S256: performing fault dislocation risk estimation on the geological fold / fault profile morphology data according to the stress progressive superposition data to obtain stress superposition fault dislocation risk data; Step S257: Based on the stress impact fold structural instability data and the stress superposition fault dislocation risk data, a stress impact structural instability analysis between adjacent rocks and soils is performed to obtain stress impact structural instability data between adjacent rocks and soils.

[0009] Preferably, step S3 comprises the following steps: Step S31: normalizing the stress impact structural instability data to obtain stress impact structural instability normalized data; Step S32: construct a geotechnical stress risk monitoring model for the stress impact structural instability normalized data based on the random forest algorithm to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0010] Preferably, step S32 includes the following steps: Step S321: dividing the stress impact structural instability normalized data into a training set and a test set to obtain a stress impact structural instability training set and a stress impact structural instability test set respectively; Step S322: performing feature decomposition processing on the stress impact structural instability training set to obtain a stress feature subspace training sequence; Step S323: performing sequence random sampling processing on the stress feature subspace training sequence to obtain a stress impact random sampling sequence; Step S325: constructing an initial geotechnical stress risk monitoring model for the stress impact random sampling sequence based on the random forest algorithm to obtain an initial geotechnical stress risk monitoring model; Step S326: Perform model testing on the initial geotechnical stress risk monitoring model through the stress impact structural instability test set to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0011] Preferably, the present invention further provides a geotechnical stress monitoring data analysis system for executing the geotechnical stress monitoring data analysis method as described above, the geotechnical stress monitoring data analysis system comprising: The immersed tube distribution difference summarization module is used to obtain the geological data of the bridge construction area on the river channel and the river channel bridge construction design data; summarize the pile driving and immersed tube distribution difference of the river channel bridge construction design data to obtain the pile driving and immersed tube distribution difference data; The stress impact structural instability analysis module is used to simulate the lateral stress deflection of pile driving according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transmission loss of the lateral stress deflection data of extrusion to obtain the effective stress transmission data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transmission data, and obtain the stress impact structural instability data between adjacent rocks and soils; The geotechnical stress risk monitoring model construction module is used to construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm, obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0012] The beneficial effect of the present invention is that by obtaining geological data and construction design data of the river bridge construction area, and summarizing the differences in the distribution of pile driving and sinking pipes in the construction design, accurate basic data can be provided for subsequent analysis. This process helps engineers identify the impact of different construction schemes on the rock and soil, and provides a basis for formulating scientific construction plans and optimizing design schemes. Through a detailed analysis of the differences in the distribution of pile driving and sinking pipes, it is possible to effectively predict the uneven stress and geological changes caused by construction, laying a good foundation for ensuring the safety of the project. By simulating the extrusion lateral stress deflection based on the difference data of the pile driving and sinking pipe distribution, it is possible to have a deep understanding of the stress deflection phenomenon caused by the pile driving process. This simulation result helps to assess whether the stress generated during the construction process will cause the instability of the surrounding rock and soil, and to promptly discover potential risk points. Further estimating the stress transfer loss of the extrusion lateral stress deflection data can accurately understand the stress transfer between the rock and soil bodies, and provide accurate data support for the subsequent stress impact structural instability analysis. The implementation of this step provides a scientific basis for stress control in construction and avoids safety problems caused by improper construction. The random forest algorithm is used to construct a geotechnical stress risk monitoring model, which provides an efficient solution for stress risk management during geotechnical engineering construction. Through machine learning analysis of stress impact structural instability data, the model can automatically identify potential risk areas and issue timely warnings, thereby realizing intelligent monitoring and accurate risk prediction. The construction of this model not only improves the accuracy and real-time performance of monitoring, but also greatly reduces the cost and error of manual monitoring, and enhances the safety during the construction process. Finally, the monitoring model is sent to the terminal, which can realize remote real-time monitoring and data analysis, and further optimize the construction control and decision-making process. Therefore, the present invention is an improved processing of a traditional geotechnical stress monitoring data analysis method, which solves the problem that a traditional geotechnical stress monitoring data analysis method has low accuracy in identifying structural instability caused by stress impact, thereby causing large errors in geotechnical stress monitoring, improves the accuracy of identifying structural instability caused by stress impact, and reduces the error of geotechnical stress monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A schematic diagram of the steps of a geotechnical stress monitoring data analysis method; Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart; Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG. DETAILED DESCRIPTION

[0014] See also Figures 1 to 3 , a geotechnical stress monitoring data analysis method, the method comprising the following steps: Step S1: obtaining geological data of the bridge construction area on the river and design data of the river bridge construction; summarizing the differences in the distribution of pile driving and immersed pipes on the river bridge construction design data to obtain data on the differences in the distribution of pile driving and immersed pipes; Step S2: simulate the lateral stress deflection of pile driving extrusion according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transfer loss of the lateral stress deflection data of extrusion to obtain the effective stress transfer data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transfer data, and obtain the stress impact structural instability data between adjacent rocks and soils; Step S3: construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm to obtain a geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0015] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a geotechnical stress monitoring data analysis method of the present invention. In this example, the geotechnical stress monitoring data analysis method includes the following steps: Step S1: obtaining geological data of the bridge construction area on the river and design data of the river bridge construction; summarizing the differences in the distribution of pile driving and immersed pipes on the river bridge construction design data to obtain data on the differences in the distribution of pile driving and immersed pipes; In the embodiment of the present invention, in order to obtain the geological data of the bridge construction area on the river and the river bridge construction design data, it is necessary to use a variety of data collection methods to conduct a detailed survey of the geological conditions of the bridge construction area. First, a high-resolution geological radar device is used to perform a three-dimensional scan of the rock and soil structure of the construction area to obtain data such as underground rock layer distribution, soil density, and water content, and the data obtained by the geological radar is further verified and supplemented in combination with the drilling exploration data. Subsequently, the density of the underground strata of the river channel is analyzed by acoustic wave detection technology to determine the characteristics of the rock layers at different depths. For the distribution area of ​​the aquifer, the resistivity tomography method is used to perform permeability detection to obtain the permeability coefficients of different rock layers. The acquired geological data is stored in the geological data database and classified and stored according to different depths and different rock layer types for subsequent analysis and use. At the same time, the construction design data of the bridge is obtained from the construction unit, including the pile driving depth, pile foundation material, sinking pipe method, sinking pipe distribution location and corresponding construction process parameters. These data are stored in the design data database after being standardized. For the distribution of pile driving and sinking pipes in the construction design data, a method based on spatial statistical analysis is used to summarize the distribution of the pile driving area. First, the pile foundation layout of the bridge construction area is spatially discretized, and the construction area is divided into multiple grid units. Each grid unit contains corresponding piling information, such as pile density, pile diameter, pile spacing and other parameters. The Kriging interpolation method is used to interpolate the piling parameters between different grid units to obtain the spatial distribution characteristics of the piling parameters in different construction areas. On this basis, the coefficient of variation of the distribution of pile sinking pipes is calculated to evaluate the uniformity of piling in different areas. For areas with large differences in pile foundation distribution, the stress interference coefficient between piles is further calculated to determine the impact of pile construction in different areas on adjacent pile foundations. Finally, the overall differences in the distribution of pile sinking pipes are summarized to form the difference data of the distribution of pile sinking pipes, which provides data support for subsequent analysis.

[0016] Step S2: simulate the lateral stress deflection of pile driving extrusion according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transfer loss of the lateral stress deflection data of extrusion to obtain the effective stress transfer data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transfer data, and obtain the stress impact structural instability data between adjacent rocks and soils; In the embodiment of the present invention, the simulation of the lateral stress deflection caused by pile driving extrusion is performed based on the distribution difference data of the pile driving pipe sinking. The stratum stress model of the pile driving construction process is constructed by the finite element analysis method. Representative geotechnical units in the construction area are selected, and the mechanical parameters of the geotechnical, such as elastic modulus, Poisson's ratio, internal friction angle, cohesion, etc., are determined according to the drilling data. In the simulation process, the lateral extrusion effect of the sinking pipe on the surrounding rock and soil during the construction process is considered, and the influence of different sinking methods (such as vibration sinking pipe and hammer sinking pipe) on the lateral stress deflection is compared. The plastic mechanics model is used to simulate the plastic deformation of the rock and soil during the pile driving process, and the propagation path and deflection angle of the lateral stress are calculated to obtain the extrusion lateral stress deflection data. The stress transmission loss of the extrusion lateral stress deflection data is estimated. Based on the stress transmission theory of layered rock and soil, a stress loss calculation model between different strata is established. According to the stress transmission efficiency of different rock strata, the elastic-plastic mechanics method is used to calculate the attenuation coefficient of the construction stress. During the analysis, the influence of discontinuity on the rock interface on stress transfer is considered, and the stress attenuation ratio between layers is calculated. For areas with abundant groundwater, the calculation of additional stress loss caused by infiltration is further introduced. After comprehensive calculation, the effective stress transfer data is obtained and stored in the database to provide input data for subsequent analysis. Based on the effective stress transfer data, the stress impact structural instability analysis between adjacent rocks and soils is carried out. A stratum contact stress analysis model is established to simulate the stress impact effect between adjacent rock and soil units. According to the bonding characteristics between different rock layers, the interface friction coefficient and shear strength are calculated, and the potential shear failure area caused by stress impact is analyzed. The stratum displacement field analysis method is used to calculate the local settlement of the stratum in the construction area, and it is evaluated whether the construction process will cause uneven settlement of the overall foundation. During the simulation process, the influence of dynamic loads on the rock and soil is considered, the seismic effect caused by the bridge construction process is analyzed, and the risk of local soil structure instability caused by dynamic loads is calculated. Finally, the stress impact structural instability data between adjacent rocks and soils are obtained, providing a basis for subsequent monitoring.

[0017] Step S3: construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm to obtain a geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0018] In an embodiment of the present invention, a random forest algorithm is used to construct a geotechnical stress risk monitoring model for stress impact structural instability data. The random forest algorithm consists of multiple decision trees, each of which is trained based on different instability characteristic parameters, and finally forms a comprehensive geotechnical stress risk assessment model. First, feature extraction is performed on the stress impact structural instability data, including stress impact intensity, stress action time, stratum deformation rate, etc. Different features are normalized using a data preprocessing method to eliminate the influence of different dimensions on the calculation results. The cross-validation method is used to optimize the parameters of the random forest model, and the K-fold cross-validation method is introduced to ensure the stability and generalization ability of the model. During the training process, the model is initially verified using historical geological data, and the model is iteratively updated using newly collected geotechnical stress monitoring data. Finally, the trained geotechnical stress risk monitoring model can predict the geotechnical stress in the construction area in real time, judge the potential risk of stratum instability, and generate corresponding monitoring and early warning data. The geotechnical stress risk monitoring model is sent to the terminal to support real-time monitoring and construction safety management.

[0019] Step S1 includes the following steps: Step S11: Acquire geological data of the bridge construction area on the river and design data of the river bridge construction; Step S12: marking the river bridge construction design data with a construction time series to obtain construction design time marked data; Step S13: performing piling sequence selection analysis on the construction design time-stamped data to obtain construction piling sequence selection data; Step S14: Summarize the differences in the distribution of pile driving and pipe immersion according to the construction piling sequence selection data and obtain the piling and pipe immersion distribution difference data.

[0020] In the embodiment of the present invention, the geological data of a construction area on the river channel and the design data of the bridge construction are first obtained. The process of obtaining geological data includes arranging exploration boreholes within the range of the longitudinal coordinate points of the river channel (120.560, 35.760) to (121.030, 35.910), drilling core samples with a depth of 50 meters along the river channel at an average interval of 200 meters, and recording the rock mass occurrence, rock layer structure and integrity information through on-site survey, and marking the hardness of different rock layers according to the pre-established rock and soil density determination standard. During the drilling exploration, the structural characteristics of each geological profile are calibrated at a resolution of 5 meters in depth, including rock particle size, mineral composition distribution and preliminary pore water content determination. When obtaining the bridge construction design data, the pile foundation position, immersed pipe process plan, pile foundation length and diameter parameters involved in the design drawings are sorted out, and the construction plan is recorded with reference to the pre-positioned coordinate information of each pile foundation in the engineering drawings. Subsequently, the construction phase information in the overall design document of the bridge was sorted out, including the start time of construction, the number of days for bridge foundation piling, and the mileage of segmented operations, so as to mark the construction sequence later. Then, the bridge construction design data was indexed according to the established time sequence. By comparing the construction phase number recorded in the design text and the date sequence, the piling work period of each week was listed in the data table, and the start and end dates of piling for each pile foundation were marked in days. Combined with the segmented operation arrangement of the bridge foundation, the corresponding piling work area number was clearly specified for each period. In this way, a construction time series table containing specific daily operation links, piling targets, and proposed immersed tube types was generated, and the precise calibration of the time series was completed in the data table by corresponding dates to piling marks. Then, based on the marked time series table, based on multiple on-site surveys and the construction pile position layout diagram provided by the design institute, a systematic analysis of the piling sequence of each period was carried out to determine the strategy to maintain stable construction efficiency between rock formation strength and river channel construction convenience. The analysis method includes an overall assessment of the number of piles driven per day, the peak load of the pile driving instrument, and the stability index of the geological structure. The rock bearing capacity and the cohesion of the surrounding riverbed soil are classified and sorted in combination with the actual geological survey results. Then, based on the evaluation results, a strategy is established to first drive areas with relatively loose geology and relatively low bearing capacity, so as to disperse the construction risks and maintain the continuous operation intensity of the pile driving equipment. On this basis, a sequence list containing the order of multiple pile driving is formed to clarify the association and allocation relationship between the early pile positions and the later pile positions. After obtaining the pile driving sequence, the sinking pipe form and pipe diameter specification of each pile foundation on a given date are extracted according to the correspondence between the sequence and the time series table. By comparing the sinking pipe type and specific length parameters of the pile foundation in the same area in the original design, the difference distribution of the pile driving and sinking pipe methods used in each construction surface, middle and deep layer on the date axis is statistically analyzed, including multi-dimensional data on the outer diameter of the sinking pipe, the wall thickness of the pipe, the number of sinking pipe sections, and the penetration force.In this process, the correspondence between the depth of the rock-soil interface and the sinking position of the pipe in the exploration records and design documents was compared. By systematically classifying the pile driving periods and geological profile characteristics of the surrounding pile foundations, a comparison table was formed for the differences in the rock-soil bearing characteristics of different depth sections and the distribution of the pipes in the corresponding construction periods. Finally, a complete analysis result of the difference in the arrangement of the pipes was obtained, covering the pile position offset, the difference in the diameter of the pipes, and the difference in the number of pipe sections in the same area during different construction periods. These difference information was summarized in the form of a data list, thus achieving the goal of summarizing the differences in the distribution of pile driving and pipes. Through the above operations, the summarized data on the difference in the distribution of pile driving and pipes can directly reflect the relationship between the construction nodes of the river bridge and the layout of the pile foundation pipes, providing an accurate and detailed pre-basis for the subsequent simulation of the extrusion lateral stress deflection.

[0021] Step S2 includes the following steps: Step S21: performing vertical geological and geotechnical structure analysis of different piling and immersion pipe areas on the geological data of the bridge construction area based on the piling and immersion pipe distribution difference data to obtain vertical geological and geotechnical structure data of different piling and immersion pipe areas; Step S22: simulating the lateral stress deflection caused by pile driving according to the pile driving and pipe sinking distribution difference data on the vertical geological and geotechnical structure data to obtain the lateral stress deflection data caused by pile driving; Step S23: identifying the stress radial influence orientation on the extrusion lateral stress deflection data to obtain the stress radial influence orientation; Step S24: estimating stress transfer loss based on the stress radial influence orientation to obtain effective stress transfer data; Step S25: Based on the effective stress transfer data, the vertical geological rock and soil structure data is subjected to stress impact structural instability analysis between adjacent rocks and soils to obtain stress impact structural instability data between adjacent rocks and soils.

[0022] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes: Step S21: performing vertical geological and geotechnical structure analysis of different piling and immersion pipe areas on the geological data of the bridge construction area based on the piling and immersion pipe distribution difference data to obtain vertical geological and geotechnical structure data of different piling and immersion pipe areas; In the embodiment of the present invention, based on the corresponding information between the difference data of pile driving and pipe sinking distribution and the geological data of the bridge construction area, the vertical geological and geotechnical structures of different pile driving and pipe sinking areas are deeply analyzed. During this implementation operation, the construction locations covered by the same type of pile driving and pipe sinking are uniformly marked on the geological map according to the coordinate range involved in the river bridge project, and further divided according to the outer diameter of the sinking pipe, the penetration depth and the distribution characteristics of the surrounding rock and soil layers. Then, according to the drilling core information obtained in the early stage, the particle grading, water content and permeability coefficient of the layered rock and soil in the area are statistically divided, and the rock layer interface and the actual influence range of the pile driving and pipe sinking are matched by layered comparison. The pipe diameter specification, pipe wall thickness and pile driving force in the difference data of pile driving and pipe sinking distribution form several constraints to distinguish the bearing capacity and density of typical vertical layers in geological data. In this process, in order to ensure the accuracy of the description of the rock and soil layer profile, multiple iterative discrete operations are used to locate the interface between the rock layer and the deposit, and the distribution of the clay layer and sandstone layer in the depth direction is corrected in combination with the measured pore water pressure data. The longitudinal profile of the geological data is subdivided into a discrete interval at a depth of every 10 meters. The lithological indicators in each interval are enhanced and compared to identify the interaction between the high water-containing sand layer and the high plasticity clay layer. The vertical geological geotechnical structure data of different pile-driving and pipe-sinking areas are generated to facilitate the subsequent fine analysis of the lateral stress deflection conditions. For example, assuming that the distribution difference data of pile-driving and pipe-sinking are: the distribution type of pile-driving and pipe-sinking is uniform distribution, and the distribution density of pile-driving and pipe-sinking is 0.5m-1. The geological data of the bridge construction area are: the rock and soil type is sand, the rock and soil density is 1.8g / cm3, and the rock and soil elastic modulus is 100MPa. The finite element analysis method is used to analyze the vertical geological geotechnical structure of different pile-driving and pipe-sinking areas of the geological data of the bridge construction area, including parameters such as rock and soil type, rock and soil density, and rock and soil elastic modulus, to obtain the vertical geological geotechnical structure data of different pile-driving and pipe-sinking areas. For example, the vertical geological and geotechnical structure data of different pile driving and pipe sinking areas are obtained as follows: the geotechnical type is sandy soil, the geotechnical density is 1.8 g / cm3, and the geotechnical elastic modulus is 100 MPa.

[0023] Step S22: simulating the lateral stress deflection caused by pile driving according to the pile driving and pipe sinking distribution difference data on the vertical geological and geotechnical structure data to obtain the lateral stress deflection data caused by pile driving; In an embodiment of the present invention, after obtaining the vertical geological and geotechnical structure data, a simulation of the lateral stress deflection caused by pile driving extrusion is performed based on the difference data of the distribution of pile driving and pipe sinking. First, according to the geotechnical parameters of different piling areas, different construction conditions such as pile foundation penetration speed, pile driving impact force, and pipe sinking depth are set. During the simulation process, the vertical load and lateral extrusion force generated by pile driving are decomposed and calculated, and the lateral stress offset characteristics of the rock and soil around the pile foundation when subjected to the extrusion force are analyzed in detail. A three-dimensional elastic-plastic constitutive model is used to describe the stress-strain relationship of the rock and soil. On this basis, the Lagrangian finite element method is used to calculate the deformation and stress change distribution of the rock and soil particles after being subjected to force, and the lateral stress deflection angle, stress diffusion range, and stress concentration area generated by the rock and soil at different piling positions are obtained. Furthermore, the stress trajectory analysis method is used to track the stress transfer path of the soil around the pile foundation to determine the attenuation of the extrusion stress in different areas and its influence on the surrounding rock and soil structure.

[0024] In another embodiment, the vertical geological geotechnical structure data is simulated for pile driving extrusion lateral stress deflection according to the difference data of pile driving and pipe sinking distribution to obtain the extrusion lateral stress deflection data. This operation first discretizes the transmission form of lateral extrusion force according to the contact parameters between the outer diameter of the pile driving and pipe sinking and the rock and soil layer, and divides the contact area between the pipe wall and the rock and soil medium into several circumferential units. Each circumferential unit is further subdivided into 3 to 5 depth slices in the longitudinal direction, corresponding to the same or similar geotechnical mechanical characteristics. Referring to the rock and soil strength index and particle density obtained by drilling, the slices of different depths are successively assigned corresponding Poisson's ratio and internal friction angle. Then, for these circumferential units and depth slices, the pile driving penetration force parameters are introduced one by one to simulate the circumferential extrusion phenomenon of the surrounding soil caused by the downward pressure of the pile foundation. In multiple iterative mixed operations, the tectonic stress field generated by each unit at a given penetration force is numerically discretized, and the stress distribution of each unit under the maximum circumferential compression state is statistically analyzed. Combined with the superposition effect of cohesion and friction resistance, the lateral deformation amplitude deflection is identified, and the lateral stress of all units is combined accordingly, and the results are recorded to form the extrusion lateral stress deflection data. During the whole process, the actual cohesion and humidity conditions of each layer of rock and soil need to be recalibrated to ensure that the lateral stress deflection results remain stable and conform to the actual rock and soil characteristics.

[0025] Step S23: identifying the stress radial influence orientation on the extrusion lateral stress deflection data to obtain the stress radial influence orientation; In an embodiment of the present invention, the radial influence direction of the extruded lateral stress deflection data is identified to obtain the radial influence direction of the stress. First, according to the lateral stress deflection result obtained in the previous step, the circumferential unit around the pile foundation is divided into sectors, and each sector is separated by an angle interval of 15 to 20 degrees to clarify the degree of stress divergence of the pile driving load in different directions. For each sector block, the magnitude and direction of the stress component are calculated in layers according to the depth reference point in the longitudinal direction, and the segmented superposition method is used to combine the penetration depth of the immersed tube and the rock thickness data. In this way, a multi-directional stress vector distribution diagram surrounding the pile foundation can be formed, and several sections with peak stress values ​​higher than the threshold are marked on the diagram. Then, through the segmented sector scanning method, according to the ratio of the peak stress value to the mean of the entire area around the pile, several radially affecting main direction angles are counted and listed. In order to avoid discrete deviation, in this process, several depth slices in each sector block are subjected to stress amplitude maximum value analysis, and the continuity of the maximum amplitude value in the same direction is retained. By integrating the obtained directional data, we can obtain the directional information of the radial distribution of stress around the pile foundation, and finally establish the radial influence direction of stress around the pile.

[0026] Step S24: estimating stress transfer loss based on the stress radial influence orientation to obtain effective stress transfer data; In the embodiment of the present invention, combined with the directional characteristics of the aforementioned radial influence orientation, the permeability coefficient, water content and geotechnical cohesion in each directional fan-shaped block are comprehensively evaluated. The specific method first applies correction coefficients in the vertical and tangential directions to the geotechnical column unit from the surface to the deepest pile point in each direction, which is used to characterize the absorption rate and dissipation rate of the geotechnical pore network to stress energy. Then, combined with the radial stress field amplitude extracted in the previous step, the energy reduction ratio caused by penetration and the energy attenuation ratio caused by friction between strata are deducted in turn, and these attenuation values ​​are compared with the reference stress amplitude to obtain the remaining stress value that can be used for transmission. The residual stress values ​​of the fan-shaped blocks in multiple directions are then integrated in the same coordinate system to form an overall effective stress distribution table, and the loss rate curves in different directions are given. By accumulating the loss rate in sections in the depth direction, the residual stress transfer values ​​corresponding to each geotechnical layer are obtained, which are generally summarized as effective stress transfer data for subsequent correlation analysis of geotechnical structures.

[0027] In another embodiment, the stress transfer loss estimation is based on the stress radial influence azimuth data, and the stress attenuation calculation is performed on the rock and soil in different regions. First, the key physical parameters such as the permeability coefficient, porosity and water content of the rock and soil at different depths are obtained. These parameters directly affect the diffusion and attenuation process of stress in the rock and soil. According to the rock and soil characteristics of different regions, the exponential decay model is used to calculate the loss rate of the extrusion stress in different directions, and the diffusion trend of the stress is analyzed by the Gaussian distribution fitting method. For the soil layer with higher permeability, the stress transfer loss is larger because its particles are loosely arranged and there is a larger porosity; while for the soil layer with higher density, the stress transfer loss is smaller. Combined with the stress path tracing technology, the stress loss between different rock and soil depths is differentially calculated to obtain the effective stress transfer data. The finite element inverse analysis method is used to verify the rationality of the calculation results to ensure that the effective stress data obtained can truly reflect the stress state of the rock and soil under the action of external forces.

[0028] Step S25: Based on the effective stress transfer data, the vertical geological rock and soil structure data is subjected to stress impact structural instability analysis between adjacent rocks and soils to obtain stress impact structural instability data between adjacent rocks and soils.

[0029] In an embodiment of the present invention, the stress impact structural instability between adjacent rocks and soils is analyzed based on the effective stress transfer data. In this process, the vertical geological rock and soil structural data are first re-layered, the interface characteristics of different geological layers are extracted, and the mechanical state of the rock and soil boundary layer is monitored. The contact mechanics analysis method is used to calculate the shear slip trend of the rock and soil boundary layer under different stresses to determine whether there is a possibility of structural instability. Combined with the principle of progressive superposition of stress, the frequency response analysis of the rock and soil stress fluctuations in different regions is carried out to identify stress concentration areas and potential risk points of stress impact. Further, the fault morphology analysis method is used to evaluate the stability of the geological fault or fold structure, and the degree of influence of stress impact on the rock and soil interface slip is calculated. Finally, the stress impact structural instability data between adjacent rocks and soils is obtained, and basic data support is provided for subsequent rock and soil stress risk monitoring.

[0030] Step S22 includes the following steps: Step S221: performing force / frequency analysis on the pile driving and immersed pipe distribution difference data to obtain the pile driving force / frequency data; Step S222: Perform vertical layered density analysis on the vertical geological rock and soil structure data to obtain the vertical layered density of the rock and soil; Step S223: Identify the lateral contact section rotation angle of the vertical layered density of the rock and soil according to the force / frequency data of the pile driving, and obtain the contact section rotation angle; Step S224: evaluating the fluctuation of the internal friction strength of the rock and soil between different layer depths of the vertical layered density of the rock and soil based on the pile driving force / frequency data and the contact section rotation angle, and obtaining the fluctuation data of the internal friction strength of the rock and soil; Step S225: performing strength boundary fluctuation integration on the rock and soil internal friction strength fluctuation data to obtain the internal friction boundary fluctuation integral value; Step S226: solving the boundary-constrained internal friction interval for the rock and soil internal friction intensity fluctuation data based on the variational method and the internal friction boundary fluctuation integral value to obtain the boundary-constrained internal friction interval; Step S227: performing a simulation of the lateral stress deflection caused by pile driving extrusion according to the friction force interval within the boundary constraint to obtain the extrusion lateral stress deflection data.

[0031] In an embodiment of the present invention, in an implementation process, firstly, based on the difference data of pile driving and sinking pipe distribution obtained in the early stage, the pile driving force and pile driving frequency of different pile foundations in each construction period are statistically analyzed to form a set of discrete sequence tables of pile driving force and frequency. To achieve this operation, the daily established pile driving process can be split to collect the number of pile driver vibrations per hour, the single penetration force, the actual action time and the variation amplitude of the penetration depth. Subsequently, the vibration recording device attached to the pile driver is used to read the continuous vibration peak value, and the peak duration is matched with the corresponding pile driving depth one by one, so as to double mark the force and frequency. In addition, by comparing the outer diameter, wall thickness grade and local rock and soil compressive strength of different types of sinking pipes, the maximum pile driving frequency interval that best meets the penetration force of the day is selected, and finally a set of pile driving force / frequency data is sorted out. After completing the analysis of force and frequency, the vertical layered density analysis is performed on the vertical geological rock and soil structure data. To achieve this goal, it is necessary to retrieve the field density measurement data of the previous drilling cores, and conduct a comprehensive comparison of the porosity, rock and soil particle arrangement and groundwater permeability coefficient of each depth section. The drilling core samples are subdivided into 5-meter or 10-meter depth sections, and the particle size distribution ratio and dry density value of each sample are measured in the laboratory. Based on the microstructural differences between adjacent depth sections, the trend of the formation density with depth is judged. In the induction, in order to highlight the stratification differences, a stratified statistical method can be used to give the average porosity and cohesion index of each layer of soil or rock layer to form a rock and soil vertical stratification density data table. Then, based on the force / frequency data applied by pile driving, the lateral contact section rotation angle under the combined action of force and frequency is identified for the rock and soil vertical stratification density. This operation requires first dividing the circumferential stress zone generated during pile driving into several discrete sectors, and the center angle of each sector can usually be set to 5 to 10 degrees to obtain higher-precision circumferential data. Then, in each sector, the pre-stated peak value of the pile vibration force is compared with the density data layer by layer, and the internal friction angle and cohesion index of the soil are used to determine the circumferential force inclination trend in the sector. When the inclination trends of all sectors are superimposed, the rotation angle distribution of a specific depth segment can be extracted, and finally the angle range of the lateral contact section deflection can be identified on the same vertical section to form a contact section rotation angle data table. Next, combined with the pile driving force / frequency data and the contact section rotation angle, the internal friction intensity between different layer depths of the vertical layered density of the rock and soil is evaluated for fluctuations. In this link, the basic friction strength value of the rock and soil corresponding to each depth layer is used as the initial reference. By introducing the tangential offset caused by the immediate impact force of the pile driving and the rotation angle, the instantaneous separation and re-overlapping behavior of the contact surface between particles in the rock and soil under stress is observed. Through a continuous discrete calculation process, the mechanical differences of particle contact in different directions are statistically analyzed, and iterative comparisons are made between adjacent depth layers to form the rock and soil internal friction intensity fluctuation data.After obtaining the internal friction intensity fluctuation data, it is necessary to perform the intensity boundary fluctuation integral on the data to obtain the internal friction boundary fluctuation integral value. This integration process relies on the layered scanning principle: in each vertical layer, the internal friction intensity fluctuation curve is recorded in the form of discrete points, and then the fluctuation values ​​between adjacent discrete points are accumulated to obtain the total interval fluctuation of each layer. These layered fluctuation totals are then summarized vertically to form the cumulative integral value of the entire depth profile. This value reflects the overall fluctuation amplitude of the rock and soil internal friction between the upper and lower layers under the piling condition, which can be used to solve the subsequent boundary constraint range. In the further processing link, the boundary constraint internal friction interval of the rock and soil internal friction intensity fluctuation data is solved by combining the aforementioned variational method theory and the internal friction boundary fluctuation integral value. Specifically, by inputting the fluctuation variance of the internal friction and the cumulative integral data information of the previous step into the variational method framework, the critical interval that produces large fluctuations is located using a layer-by-layer iterative discrete method. In each depth segment, the fluctuation peak and valley values ​​are selected as the segmentation points, and the upper and lower limits of the internal friction of the rock and soil medium within the layer range are determined by combining the contact section rotation angle information obtained by the previous processing. The upper and lower limits in each layer are merged to form the final boundary-constrained internal friction interval. This interval can intuitively show the bearing limit of rock and soil at different depths for pile lateral stress. Finally, based on the boundary-constrained internal friction interval, the comprehensive numerical calculation of pile extrusion lateral stress deflection is performed on the rock and soil medium on the vertical geological section, thereby obtaining the extrusion lateral stress deflection data. At this stage, the extrusion force can be first applied to the discrete area around the pile foundation, and the upper and lower limits of the internal friction force obtained previously are included in the calculation to correct the rebound or plastic deformation of each layered soil body for lateral stress. When performing the analysis for each discrete cell, it is necessary to consider the internal friction difference of the surrounding cells and the rotation angle of the contact section at the same time, and introduce the layered accumulation method to record the deflection increment. After the data of all cells are converged, the overall distribution result of pile extrusion lateral stress deflection can be found in the three-dimensional space of the entire pile area. This result provides comprehensive data support for the identification of the radial influence orientation of the subsequent stress field and the estimation of stress loss.

[0032] Step S24 includes the following steps: Step S241: obtaining the moisture content at different rock and soil depths; mapping the moisture content at different rock and soil depths in terms of stress diffusion orientation based on the stress radial influence orientation to obtain the moisture content at the stress diffusion orientation; Step S242: evaluating the water pressure increment trend between porous rock and soil particle mixtures based on the moisture content in the stress diffusion direction and the moisture content at different rock and soil depths, and obtaining rock and soil particle mixture water pressure increment data; Step S243: evaluating the contact kinetic energy damping ratio of the rock-soil particle mixture water pressure increment data to obtain the particle mixture water pressure kinetic energy damping ratio; Step S244: Estimate the stress transfer loss of the extrusion lateral stress deflection data based on the particle mixture water pressure kinetic energy damping ratio to obtain effective stress transfer data.

[0033] In the embodiment of the present invention, in the process of obtaining the moisture content at different rock and soil depths, it is necessary to use drilling sampling and moisture sensing technology for joint measurement. First, multiple depth sensing points are arranged in the construction area, and rock and soil samples at different depths are drilled using small-caliber drilling equipment. The drilling depth is controlled within 20 meters underground, and sampling points are arranged at intervals of 1 meter. During the sampling process, the natural moisture content of the original soil sample is determined by the rapid weighing method, and the soil sample is dried at a constant temperature of 105°C by the drying method, and the mass loss is calculated to obtain the moisture content data. For fine-grained soil, a high-frequency electromagnetic moisture sensor is used to measure the conductivity, and the moisture content is calculated in combination with the soil type correction formula. During the data recording process, the moisture content data at different depths are stored according to the spatial coordinates, and a deep moisture distribution database is established. In order to ensure the accuracy of the data, each measuring point is measured at least three times, and the mean of the measurement error is calculated, and the abnormal data whose error exceeds the set threshold is eliminated. In the process of stress diffusion azimuth moisture content mapping, it is necessary to calculate the influence area of ​​moisture distribution based on the radial diffusion characteristics of rock and soil stress. First, based on the geological survey data of the construction area, the stress diffusion radius of the rock and soil layer is determined, and the stress diffusion gradient in different directions is calculated according to the soil density and elastic modulus. The capillary water effect between rock and soil particles at different depths is calculated using the moisture migration equation, and the dynamic distribution of moisture under stress is estimated by combining the empirical formula. In the data processing process, the moisture content of unmeasured points is calculated using the spatial interpolation method, and the measured data is smoothed using the weighted average algorithm to eliminate local outliers. Finally, a stress diffusion azimuth moisture content database is established based on the calculation results, and the data is associated with the rock and soil depth coordinates to form a complete spatial mapping relationship. In the process of evaluating the trend of the mixed water pressure increment of rock and soil particles, it is necessary to combine the stress diffusion azimuth moisture content with the moisture content at different depths to calculate the change trend of pore water pressure. First, the moisture content data at each depth in the measurement area is extracted, and the seepage velocity of pore water is calculated based on the soil permeability coefficient. The volume change of pore water under stress is calculated using the Biot consolidation theory, and the water pressure increment is solved by the numerical difference method. For saturated soil layers, the water permeability is calculated in combination with Darcy's law, and the water pressure growth trend is solved by the time step recursion method. During the data storage process, a three-dimensional water pressure increment distribution map is constructed according to the spatial coordinates of the construction area, and the curve fitting method is used to analyze the trend of water pressure over time. Finally, all calculation results are archived in the rock and soil particle mixed water pressure increment database and used for subsequent stress analysis. In the process of evaluating the kinetic energy damping ratio of the particle mixed water pressure, it is necessary to calculate the kinetic energy loss of pore water based on the rock and soil particle mixed water pressure increment data. First, the water pressure increment data at different depths in the construction area are extracted, and the initial kinetic energy of water movement is calculated in combination with the fluid dynamics equation. The Navier-Stokes equation is used to calculate the flow resistance of pore water between rock and soil particles, and the decay rate of water pressure kinetic energy is solved by the numerical integration method.For soil layers with different particle gradations, the particle contact model is used to calculate the energy transfer efficiency between particles, and the nonlinear fitting method is used to analyze the change trend of the damping ratio. In the data processing process, the water dynamic loss rate is calculated based on the particle diameter distribution of rock and soil particles, and cross-validated with the foundation stress data. Finally, all calculation results are stored in the particle mixed water pressure kinetic energy damping ratio database and used for stress transfer loss estimation.

[0034] In the process of stress transfer loss estimation, it is necessary to calculate the effective stress transfer of the extrusion lateral stress deflection data based on the particle mixed water pressure kinetic energy damping ratio. First, the damping ratio data in the construction area is extracted, and the attenuation trend of the lateral stress is calculated in combination with the foundation stress. The contact force chain analysis method is used to calculate the stress transfer efficiency between geotechnical particles, and the stress loss rate at different depths is solved by combining the numerical integration method. For soil layers with high moisture content, the modified Terzaghi effective stress formula is used to calculate the influence of pore water pressure, and the deflection path of the extrusion lateral stress is evaluated in combination with the dynamic stress analysis method. In the data storage process, a spatial database is constructed according to the stress loss distribution at different depths, and the stress loss contour map is generated using data visualization technology. Finally, all calculated data are stored in the effective stress transfer database and used for subsequent geotechnical stress monitoring and analysis.

[0035] Step S25 includes the following steps: Step S251: performing geotechnical classification on the vertical geological and geotechnical structure data to obtain vertical geological and geotechnical classification types; Step S252: Based on the vertical geological and rock-soil classification type, the vertical geological and rock-soil structure data is subjected to geological fold / fault profile morphology recognition between adjacent rocks and soils to obtain geological fold / fault profile morphology data; Step S253: performing stress progressive superposition frequency evaluation on the effective stress transfer data to obtain stress progressive superposition data; performing fold deformation inclination angle calculation on the geological fold / fault profile morphology data to obtain the fold deformation inclination angle; Step S254: calculating the average difference of fold surface slopes at different depths for the geological fold / fault profile morphology data according to the fold deformation inclination angle, and obtaining the average difference of fold surface slopes; Step S255: Based on the stress progressive superposition data, a stress impact structural instability simulation is performed on the fold deformation inclination angle and the fold surface slope difference to obtain stress impact fold structural instability data; Step S256: performing fault dislocation risk estimation on the geological fold / fault profile morphology data according to the stress progressive superposition data to obtain stress superposition fault dislocation risk data; Step S257: Based on the stress impact fold structural instability data and the stress superposition fault dislocation risk data, a stress impact structural instability analysis between adjacent rocks and soils is performed to obtain stress impact structural instability data between adjacent rocks and soils.

[0036] In an embodiment of the present invention, a description of the core is obtained from the vertical geological rock and soil structure data, including characteristics such as color, particle size, mineral composition, degree of cementation, and porosity. In the process of geotechnical type classification of vertical geological rock and soil structure data, it is also necessary to comprehensively process the drilling exploration data, core analysis results, and geophysical logging data. First, multiple drilling points are selected in the construction area, and the core is described layer by layer, including characteristics such as color, particle size, mineral composition, degree of cementation, and porosity. X-ray diffraction analysis is used to quantitatively determine the mineral composition of the rock sample, and mechanical parameters such as compressive strength, elastic modulus, and Poisson's ratio are determined in combination with rock mechanics tests. The resistivity of rock and soil layers at different depths is measured by tomographic resistivity imaging, and the rock and soil density and structural characteristics are calculated in combination with seismic wave velocity inversion. Based on the geotechnical mechanics classification standard, the rock and soil structure is divided into different types such as sand, silty clay, bedrock, etc., and the division results are cross-validated to finally form a vertical geological rock and soil classification type database. In the process of identifying the morphology of geological folds / fault profiles based on vertical geological geotechnical structural data, it is necessary to analyze the geometric morphology and deformation characteristics of adjacent rock layers based on geotechnical type data. First, the continuity of underground rock layers is extracted using high-precision seismic profile data, and the geological interface characteristics are identified through a multi-scale edge detection algorithm. For the curved morphology of the rock layer, the curvature analysis method is used to calculate the position of the fold axis, and the position of the fracture surface is determined by combining the fault dip analysis. For fault development areas, the fault dislocation amplitude is calculated based on the dislocation gradient, and the fault type is determined by displacement vector analysis. The terrain linear feature extraction method is used to identify the fault trace, and the fault activity is evaluated in combination with historical seismic data. Finally, all calculation results are stored in the geological fold / fault profile morphology database for subsequent stress analysis. In the process of stress progressive superposition frequency evaluation, effective stress transfer data is needed to calculate the stress accumulation trend of rock and soil bodies at different depths. First, the ground stress data of the study area is extracted, and the stress gradient change is calculated using the differential method. Based on the dynamic change data of the groundwater level, the influence of pore water pressure on stress accumulation is calculated, and periodic stress analysis is performed in combination with historical monitoring data. The Fourier transform method is used to decompose the stress change signal in the frequency domain, and the short-time Fourier transform is used to calculate the frequency of stress change over time. For different geological units, the wavelet analysis method is used to extract the local stress peak value and calculate the progressive superposition trend of the stress change cycle. All calculation results are stored in the stress progressive superposition database and used for stress impact structural instability simulation. In the process of calculating the inclination angle of fold deformation, it is necessary to use the geological fold / fault profile morphological data to calculate the inclination angle of the fold axis. First, the fold axis trace is extracted based on the geological profile data, and the least squares fitting method is used to calculate the fold curvature change rate. The numerical differentiation method is used to calculate the inclination rate of the fold axis at different depths, and the fold inclination angle is determined in combination with the rock formation deformation characteristics.For nonlinear deformation folds, the Bézier curve fitting method is used to calculate the local tilt change, and the depth distribution of the tilt angle is analyzed in combination with the rock layer thickness data. All calculation results are stored in the fold deformation tilt angle database and used for the subsequent fold surface slope mean difference calculation. In the process of calculating the fold surface slope mean difference, it is necessary to calculate the fold surface slope change trend between different depths based on the fold deformation tilt angle data. First, the fold profile data is extracted, and the slope change rate of adjacent depths is calculated using the finite difference method. For complex fold structures, the bicubic interpolation method is used to calculate the local slope difference in the non-uniform deformation area. Combined with the fold axis trace data, the mean square error of the fold surface slope is calculated, and the spatial statistical analysis method is used to evaluate the stability of the slope change. All calculation results are stored in the fold surface slope mean difference database and used for stress impact structural instability simulation. In the process of stress impact structural instability simulation, it is necessary to analyze the structural stability of the rock mass under external force based on the stress progressive superposition data, the fold deformation tilt angle and the fold surface slope mean difference. First, stress data at different depths are extracted, and the stress wave equation is used to calculate the stress impact effect. Combined with the fold deformation inclination angle data, the deformation mode of the fold structure under the impact load is analyzed. The discrete element method is used to simulate the stress transfer between rock and soil particles, and the critical point of instability is calculated in combination with the friction slip model. For high stress areas, the local stress release method is used to simulate the instability failure mode of rock formations, and the crack expansion rate is calculated in combination with the fracture mechanics method. All calculation results are stored in the stress impact fold structure instability database and used for fault dislocation risk estimation. In the process of fault dislocation risk estimation, it is necessary to evaluate the fault slip risk of the geological fold / fault profile morphology data based on the stress progressive superposition data. First, the fault profile data is extracted, and the critical stress of fault slip is calculated using the Coulomb friction criterion. Combined with the seismic wave propagation data, the dislocation trend of the fault under different loads is analyzed. The three-dimensional finite element method is used to simulate the fault dislocation displacement field, and the fault slip risk level is calculated in combination with the stress concentration factor. For active faults, a dynamic friction model is used to analyze the energy dissipation during sliding, and the fault stability is evaluated in combination with historical earthquake data. All calculation results are stored in the stress superposition fault dislocation risk database and used for stress impact structural instability analysis between adjacent rock and soil. In the process of stress impact structural instability analysis between adjacent rock and soil, it is necessary to combine the stress impact fold structure instability data and the stress superposition fault dislocation risk data to calculate the overall stability of adjacent rock and soil. First, the stress impact data of different geological units are extracted, and the stress transfer effect is calculated using the numerical integration method. Combined with the fault dislocation risk data, the impact of the fault on the adjacent rock and soil structure is analyzed, and the contact mechanics method is used to simulate the disturbance effect of fault slip on the stress distribution. The energy balance analysis method is used to calculate the possibility of instability of the rock and soil under stress impact, and the dynamic analysis method is combined to evaluate the risk of rock slip.Finally, all calculation results are stored in the stress impact structural instability database between adjacent rocks and soils, and used for geotechnical engineering safety monitoring and prediction analysis.

[0037] Step S256 includes the following steps: Decompose the stress action direction vector of the stress progressive superposition data to obtain the stress data of all directions; According to the stress data of various components, the normal and tangential stress components of the fault section are calculated for the geological fold / fault section morphology data to obtain the stress component data of the fault surface; Perform fault joint distribution density analysis on geological fold / fault profile morphology data to obtain fault joint distribution density data; According to the fault plane stress component data, the critical bearing stress of the joint surface is simulated for the fault joint distribution density data to obtain the critical bearing stress of the joint surface; The fault dislocation risk is estimated according to the critical bearing stress of the joint surface, and the stress superimposed fault dislocation risk data is obtained.

[0038] In an embodiment of the present invention, the stress progressive superposition data is decomposed by the stress action direction vector to obtain the component stress data in all directions. In the process of geotechnical stress monitoring data analysis, the stored stress progressive superposition data is first extracted, and the three-dimensional stress tensor is decomposed into the main stress components by using the tensor decomposition method. Based on the Mohr stress circle theory, the stress tensor is transformed by the principal stress axis rotation, the maximum, minimum and intermediate principal stress components are calculated, and the principal stress is further decomposed into the local coordinate system by using the coordinate transformation matrix to obtain the component stress data along a specific direction. The spatial gradient of the stress field is calculated by the finite difference method to determine the directionality of the stress change at different positions, and the continuity of the stress field is analyzed by combining the Gaussian curvature. The grid units are divided in the stress field area, and the generalized stress formula is used for each unit to calculate the normal stress and shear stress acting in different directions to form a component stress data set in all directions. Combined with the distribution characteristics of the stress field, the interpolation algorithm is used to smooth the data to reduce the impact of the measurement error and ensure the stability and accuracy of the component stress data in all directions. According to the various component stress data, the normal and tangential stress components of the fault profile are calculated for the geological fold / fault profile morphology data to obtain the fault surface stress component data. According to the fault profile morphology data, the normal vector of the fault surface is extracted, and the normal stress and tangential stress on the fault surface are calculated in combination with the various component stress data. First, the normal direction of the fault surface in the three-dimensional coordinate system is calculated using the spatial geometric projection method, and a local coordinate system is established to facilitate the projection calculation of the stress components. Based on the Cauchy stress formula, the stress components acting on the fault surface are calculated, where the normal stress is obtained by the dot product of the stress tensor and the fault normal vector, and the tangential stress is calculated by the dot product of the stress tensor and the fault tangent vector. The local differential method is used to analyze the spatial variation of the fault surface stress, determine the stress concentration area, and combine the Fourier transform to analyze the distribution characteristics of the stress components at different scales. For different fault types, the ratio of shear stress to normal stress is calculated according to the shear rupture criterion to determine the fault surface rupture mode. All calculation results are stored in the fault surface stress component data set and used for subsequent fault stability analysis. The fault joint distribution density is analyzed on the geological fold / fault profile morphological data to obtain the fault joint distribution density data. Based on the fault profile morphological data, the spatial distribution information of the joints near the fault is extracted, and the image processing algorithm is used to detect the edge of the core scanning image, identify the joint structure, and calculate the joint spacing and density. Based on the Hough transform method, the spatial trend of the joints is identified, and the distribution characteristics of the joints are determined in combination with the fault profile data. The Kriging interpolation method is used to perform spatial statistical analysis on the joint distribution, calculate the trend of joint density changes in different regions, and analyze the tendency of the joints in combination with the fault plane dip data. For high-density joint areas, the Voronoi diagram segmentation method is used to divide the joint units, and the joint density distribution of each unit is calculated.Combined with the characteristics of fault slip, the fractal dimension analysis method is used to evaluate the complexity of joints and calculate their impact on fault stability. All calculation results are stored in the fault joint distribution density data set and used for subsequent fault slip risk assessment. According to the fault surface stress component data, the critical bearing stress of the joint surface is simulated for the fault joint distribution density data to obtain the critical bearing stress of the joint surface. Based on the joint distribution density data and the fault surface stress component data, the stress equilibrium equation on the joint surface is established, and the bearing stress in different joint directions is calculated. The shear strength of the joint surface is calculated according to the Mohr-Coulomb criterion using the numerical iteration method, and the bearing capacity of the joint surface is calculated in combination with the normal stress of the joint surface. For joints in different directions, the rigid-plastic finite element method is used to simulate the stress concentration phenomenon of the joint surface, and the critical shear stress is calculated. For areas with small joint spacing, the particle flow method is used to simulate the stress transfer process of the joint surface under the action of external force, and the critical value of the failure of the stress concentration area is calculated. Combined with the spatial distribution characteristics of fault joints, the critical bearing stress of joints at different positions is calculated, and the calculation results are fitted by statistical regression method to obtain the overall distribution trend of the bearing capacity of the joint surface. All calculation results are stored in the critical bearing stress data set of joints and used for fault dislocation risk assessment. The fault dislocation risk is estimated according to the critical bearing stress of joints, and the stress superposition fault dislocation risk data is obtained. According to the critical bearing stress data of joints, the comprehensive failure probability on the fault plane is calculated, and the fault dislocation statistical model is used to analyze the possibility of fault slip. The displacement change of the fault under different loads is calculated by using the earthquake dynamics analysis method, and the slip trend is analyzed by combining the fault sliding friction model. The finite element method is used to simulate the dislocation characteristics of the fault under different stresses, and the dislocation rate and fault activity intensity are calculated. For high dislocation risk areas, regression analysis is performed in combination with historical earthquake data to predict the future activity trend of the fault. For complex fault systems, the Markov chain Monte Carlo method is used to simulate the fault slip process and calculate the probability distribution of fault slip risk. All calculation results are stored in the stress superposition fault dislocation risk data set and used for stability assessment and early warning analysis of geotechnical engineering.

[0039] Step S3 includes the following steps: Step S31: normalizing the stress impact structural instability data to obtain stress impact structural instability normalized data; Step S32: construct a geotechnical stress risk monitoring model for the stress impact structural instability normalized data based on the random forest algorithm to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0040] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes: Step S31: normalizing the stress impact structural instability data to obtain stress impact structural instability normalized data; In the embodiment of the present invention, the stress impact structural instability data is normalized to obtain stress impact structural instability normalized data. First, the existing stress impact structural instability data set is extracted, including multi-dimensional stress monitoring data and related information of the instability event. These data usually contain stress values ​​at multiple different time nodes, as well as structural instability marks. For these data, the range of the maximum and minimum values ​​is first determined to avoid excessive impact of extreme values ​​in the data on subsequent processing. The data is standardized by a linear normalization method, and the specific method is to convert it by the following formula: normalized data = (original data-minimum value) / (maximum-minimum value). This method ensures that all data points are mapped to the interval of 0 to 1, so that data of different dimensions are comparable and the influence of dimension is eliminated. For missing data or outliers, interpolation is used to fill in to ensure the integrity of the data. After normalization, data smoothing is further performed, and the data is smoothed by a sliding window method so that the data has better continuity and trend in the time series. The processed normalized data set can effectively eliminate the deviation caused by the range difference of the original data and provide more consistent input data for subsequent model training and analysis.

[0041] Step S32: construct a geotechnical stress risk monitoring model for the stress impact structural instability normalized data based on the random forest algorithm to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0042] In the embodiment of the present invention, a geotechnical stress risk monitoring model is constructed based on the normalized data of stress impact structural instability based on the random forest algorithm to obtain a geotechnical stress risk monitoring model, and the geotechnical stress risk monitoring model is sent to the terminal to perform geotechnical stress monitoring data analysis. First, relevant features are extracted from the normalized stress impact structural instability data, including stress change rate, stress values ​​before and after the instability time point, stress oscillation amplitude and other multi-dimensional data, and a feature matrix is ​​constructed. The feature matrix is ​​combined with the corresponding instability label (such as instability or no instability) to construct a training data set for supervised learning. The training data set is trained using the random forest algorithm, which improves the accuracy and robustness of the model by constructing multiple decision trees and making integrated decisions. During the training process, the random forest algorithm will select features according to the different importance of the features through indicators such as the Gini index or information gain, and automatically screen out the features with the most predictive ability. Hyperparameter adjustment during the training process, such as the number of trees, maximum depth, minimum number of sample splits, etc., is optimized by the cross-validation method to ensure the generalization ability and prediction accuracy of the model. After the training is completed, the constructed random forest model will be evaluated, and the performance of the model will be tested using the test set to ensure that the model performs stably and reliably on unknown data. Finally, the optimized geotechnical stress risk monitoring model will be sent to the terminal device via a wireless network or other communication means, so that it can receive stress monitoring data from the site in real time, and conduct risk assessment on real-time data through the model, thereby providing prediction and early warning information on stress changes. The deployment and operation of the model can be automatically executed on the terminal device without manual intervention, which greatly improves the efficiency and accuracy of geotechnical stress monitoring.

[0043] Step S32 includes the following steps: Step S321: dividing the stress impact structural instability normalized data into a training set and a test set to obtain a stress impact structural instability training set and a stress impact structural instability test set respectively; Step S322: performing feature decomposition processing on the stress impact structural instability training set to obtain a stress feature subspace training sequence; Step S323: performing sequence random sampling processing on the stress feature subspace training sequence to obtain a stress impact random sampling sequence; Step S325: constructing an initial geotechnical stress risk monitoring model for the stress impact random sampling sequence based on the random forest algorithm to obtain an initial geotechnical stress risk monitoring model; Step S326: Perform model testing on the initial geotechnical stress risk monitoring model through the stress impact structural instability test set to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0044] In the embodiment of the present invention, the stress impact structural instability normalized data is divided into a training set and a test set to obtain a stress impact structural instability training set and a stress impact structural instability test set. First, a certain proportion of data is randomly extracted from the stress impact structural instability normalized data set as a training set and a test set. In order to ensure the representativeness of the training set and the test set, the stress data will be randomly divided according to the time sequence to avoid the deviation caused by the time correlation of the data set. In the division process, a ratio of 7:3 is adopted, that is, 70% of the data is used as the training set and 30% of the data is used as the test set. This ratio can ensure that the training set has enough data for model learning, and the test set can effectively evaluate the performance of the model. When dividing, the stress impact data is first deduplicated and cleaned, and outliers and duplicate data are eliminated to ensure the quality of the data. After division, the statistical characteristics of the training set and the test set are checked to ensure that the two have similar distribution characteristics in terms of stress fluctuations, instability states, etc., to ensure the reliability of subsequent analysis. The stress impact structural instability training set is subjected to feature decomposition processing to obtain a stress feature subspace training sequence. After obtaining the training set data, the data is first subjected to principal component analysis (PCA) for feature decomposition. PCA maps the original high-dimensional data to a low-dimensional space by finding the direction with the largest variance in the data to reduce redundant information and retain the most representative features in the data. First, the covariance matrix of the training set is calculated, and then the variance contribution of each feature is obtained by eigenvalue decomposition. Then, the features with larger variance contribution are selected as new feature subspaces to form a new feature matrix. Through this process, the dimension of the training set can be significantly reduced while maintaining the core information of the data, ensuring the computational efficiency and accuracy of subsequent model training. The obtained stress feature subspace training sequence contains important stress features in the training set, such as stress change rate, stress peak value, and stress value at the moment of instability. The stress feature subspace training sequence is subjected to sequence random sampling to obtain a stress shock random sampling sequence. The decomposed feature subspace training sequence is randomly sampled to simulate different stress shock scenarios. In the sampling process, the bootstrap method is used to extract each training sample with replacement. Through the bootstrap method, multiple different training data sets can be generated. These data sets have high diversity in sample distribution, which helps to improve the robustness of the model. Specifically, a certain number of data points are randomly selected from the training sequence each time to generate a new sub-sample set until the number of samples is the same as the original training set. After each sampling, the calculated stress shock random sampling sequence will be used as a new training input for further learning of the model. In this way, the diversity of training data is enhanced, thereby reducing the risk of model overfitting and improving generalization ability. Based on the random forest algorithm, the initial geotechnical stress risk monitoring model is constructed for the stress shock random sampling sequence to obtain the initial geotechnical stress risk monitoring model.After obtaining the random sampling sequence of stress impact, the random forest algorithm is used to model the data. Random forest is an integrated learning method that trains multiple decision trees and predicts by majority voting. In the modeling process, multiple decision trees are first generated according to the sampled data set. Each tree uses a different subset of the data set during the training process, and each tree adopts a random feature selection strategy when dividing the nodes to further enhance the diversity of the model. Each decision tree is continuously split according to the node selection criteria (such as information gain, Gini index) until the maximum tree depth or other termination conditions are reached. During the training process, the number of optimal trees is selected by cross-validation, and the training effect of each tree is evaluated. Finally, the output results of all decision trees are voted, the classification results of each tree are integrated, and the final stress risk prediction results are output. This initially constructed model is used for subsequent geotechnical stress risk monitoring and provides a basis for the final risk warning. The initial geotechnical stress risk monitoring model is tested by the stress impact structural instability test set to obtain the geotechnical stress risk monitoring model, and the geotechnical stress risk monitoring model is sent to the terminal to perform geotechnical stress monitoring data analysis. After the initial model construction is completed, the model is verified using the test set. The data in the test set is kept independent of the data in the training set to ensure the generalization ability of the model. By inputting the data in the test set into the constructed random forest model, the prediction results of the model are calculated and compared with the actual instability events to evaluate the accuracy of the model. The evaluation indicators include classification accuracy, precision, recall rate, F1-score, etc. Through the model testing process, the deficiencies of the model can be identified and corresponding optimization adjustments can be made. Finally, the optimized geotechnical stress risk monitoring model will be sent to the on-site monitoring terminal through communication protocols (such as wireless transmission, Bluetooth, etc.). The terminal receives the model and analyzes it based on the real-time collected stress data to provide real-time stress change monitoring and risk warning information for engineering personnel.

[0045] The present invention also provides a geotechnical stress monitoring data analysis system for executing the geotechnical stress monitoring data analysis method as described above. The geotechnical stress monitoring data analysis system comprises: The immersed tube distribution difference summarization module is used to obtain the geological data of the bridge construction area on the river channel and the river channel bridge construction design data; summarize the pile driving and immersed tube distribution difference of the river channel bridge construction design data to obtain the pile driving and immersed tube distribution difference data; The stress impact structural instability analysis module is used to simulate the lateral stress deflection of pile driving according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transmission loss of the lateral stress deflection data of extrusion to obtain the effective stress transmission data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transmission data, and obtain the stress impact structural instability data between adjacent rocks and soils; The geotechnical stress risk monitoring model construction module is used to construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm, obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

[0046] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A geotechnical stress monitoring data analysis method, characterized in that: The following steps are involved: Step S1: obtaining geological data of the bridge construction area on the river and design data of the river bridge construction; summarizing the differences in the distribution of pile driving and immersed pipes on the river bridge construction design data to obtain data on the differences in the distribution of pile driving and immersed pipes; Step S2: simulating the lateral stress deflection caused by pile driving according to the distribution difference data of pile driving and immersed pipe, and obtaining the lateral stress deflection data of pile driving; The stress transfer loss is estimated based on the extrusion lateral stress deflection data to obtain the effective stress transfer data; based on the effective stress transfer data, the stress impact structural instability analysis between adjacent rocks and soils is performed to obtain the stress impact structural instability data between adjacent rocks and soils; Step S3: construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm to obtain a geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

2. The geotechnical stress monitoring data analysis method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire geological data of the bridge construction area on the river and design data of the river bridge construction; Step S12: marking the river bridge construction design data with a construction time series to obtain construction design time marked data; Step S13: performing piling sequence selection analysis on the construction design time-stamped data to obtain construction piling sequence selection data; Step S14: Summarize the differences in the distribution of pile driving and pipe immersion according to the construction piling sequence selection data and obtain the piling and pipe immersion distribution difference data.

3. The geotechnical stress monitoring data analysis method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing vertical geological and geotechnical structure analysis of different piling and immersion pipe areas on the geological data of the bridge construction area based on the piling and immersion pipe distribution difference data to obtain vertical geological and geotechnical structure data of different piling and immersion pipe areas; Step S22: simulating the lateral stress deflection caused by pile driving according to the pile driving and pipe sinking distribution difference data on the vertical geological and geotechnical structure data to obtain the lateral stress deflection data caused by pile driving; Step S23: identifying the stress radial influence orientation on the extrusion lateral stress deflection data to obtain the stress radial influence orientation; Step S24: estimating stress transfer loss based on the stress radial influence orientation to obtain effective stress transfer data; Step S25: Based on the effective stress transfer data, the vertical geological rock and soil structure data is subjected to stress impact structural instability analysis between adjacent rocks and soils to obtain stress impact structural instability data between adjacent rocks and soils.

4. The geotechnical stress monitoring data analysis method according to claim 3 is characterized in that: Step S22 includes the following steps: Step S221: performing force / frequency analysis on the pile driving and immersed pipe distribution difference data to obtain the pile driving force / frequency data; Step S222: Perform vertical layered density analysis on the vertical geological rock and soil structure data to obtain the vertical layered density of the rock and soil; Step S223: Identify the lateral contact section rotation angle of the vertical layered density of the rock and soil according to the force / frequency data of the pile driving, and obtain the contact section rotation angle; Step S224: evaluating the fluctuation of the internal friction strength of the rock and soil between different layer depths of the vertical layered density of the rock and soil based on the pile driving force / frequency data and the contact section rotation angle, and obtaining the fluctuation data of the internal friction strength of the rock and soil; Step S225: performing strength boundary fluctuation integration on the rock and soil internal friction strength fluctuation data to obtain the internal friction boundary fluctuation integral value; Step S226: solving the boundary-constrained internal friction interval for the rock and soil internal friction intensity fluctuation data based on the variational method and the internal friction boundary fluctuation integral value to obtain the boundary-constrained internal friction interval; Step S227: performing a simulation of the lateral stress deflection caused by pile driving extrusion according to the friction force interval within the boundary constraint to obtain the extrusion lateral stress deflection data.

5. The geotechnical stress monitoring data analysis method according to claim 3, characterized in that: Step S25 includes the following steps: Step S251: performing geotechnical classification on the vertical geological and geotechnical structure data to obtain vertical geological and geotechnical classification types; Step S252: Based on the vertical geological and rock-soil classification type, the vertical geological and rock-soil structure data is subjected to geological fold / fault profile morphology recognition between adjacent rocks and soils to obtain geological fold / fault profile morphology data; Step S253: performing stress progressive superposition frequency evaluation on the effective stress transfer data to obtain stress progressive superposition data; performing fold deformation inclination angle calculation on the geological fold / fault profile morphology data to obtain the fold deformation inclination angle; Step S254: calculating the average difference of fold surface slopes at different depths for the geological fold / fault profile morphology data according to the fold deformation inclination angle, and obtaining the average difference of fold surface slopes; Step S255: Based on the stress progressive superposition data, a stress impact structural instability simulation is performed on the fold deformation inclination angle and the fold surface slope difference to obtain stress impact fold structural instability data; Step S256: performing fault dislocation risk estimation on the geological fold / fault profile morphology data according to the stress progressive superposition data to obtain stress superposition fault dislocation risk data; Step S257: Based on the stress impact fold structural instability data and the stress superposition fault dislocation risk data, a stress impact structural instability analysis between adjacent rocks and soils is performed to obtain stress impact structural instability data between adjacent rocks and soils.

6. The geotechnical stress monitoring data analysis method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: normalizing the stress impact structural instability data to obtain stress impact structural instability normalized data; Step S32: construct a geotechnical stress risk monitoring model for the stress impact structural instability normalized data based on the random forest algorithm to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

7. The method for analyzing geotechnical stress monitoring data according to claim 6, characterized in that: Step S32 includes the following steps: Step S321: dividing the stress impact structural instability normalized data into a training set and a test set to obtain a stress impact structural instability training set and a stress impact structural instability test set respectively; Step S322: performing feature decomposition processing on the stress impact structural instability training set to obtain a stress feature subspace training sequence; Step S323: performing sequence random sampling processing on the stress feature subspace training sequence to obtain a stress impact random sampling sequence; Step S325: constructing an initial geotechnical stress risk monitoring model for the stress impact random sampling sequence based on the random forest algorithm to obtain an initial geotechnical stress risk monitoring model; Step S326: Perform model testing on the initial geotechnical stress risk monitoring model through the stress impact structural instability test set to obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

8. A geotechnical stress monitoring data analysis system, characterized in that: Used to execute the geotechnical stress monitoring data analysis method according to claim 1, the geotechnical stress monitoring data analysis system comprises: The immersed tube distribution difference summarization module is used to obtain the geological data of the bridge construction area on the river channel and the river channel bridge construction design data; summarize the pile driving and immersed tube distribution difference of the river channel bridge construction design data to obtain the pile driving and immersed tube distribution difference data; The stress impact structural instability analysis module is used to simulate the lateral stress deflection of pile driving according to the distribution difference data of pile driving and immersed pipe, and obtain the lateral stress deflection data of extrusion; estimate the stress transmission loss of the lateral stress deflection data of extrusion to obtain the effective stress transmission data; perform stress impact structural instability analysis between adjacent rocks and soils based on the effective stress transmission data, and obtain the stress impact structural instability data between adjacent rocks and soils; The geotechnical stress risk monitoring model construction module is used to construct a geotechnical stress risk monitoring model for stress impact structural instability data based on the random forest algorithm, obtain the geotechnical stress risk monitoring model, and send the geotechnical stress risk monitoring model to the terminal to perform geotechnical stress monitoring data analysis.

Images