Working condition inversion system and method under ultra-deep foundation pit based on algorithm model

Abstract

The invention relates to the technical field of constructional engineering, and discloses a working condition inversion system and method under an ultra-deep foundation pit based on an algorithm model, and the system comprises a data collection module, a geological data processing module and the like. The data acquisition module collects geological and construction monitoring data, the geological data processing module evaluates geological stability, the monitoring data comparison module judges whether the monitoring data are normal or not, the model operation module obtains a working condition inversion result by means of a specific algorithm model, the inversion result verification module verifies the result, and the abnormity processing module processes abnormity. According to the method, data can be accurately collected and analyzed, working conditions can be evaluated in real time, abnormity can be early warned in time, the inversion precision is improved by utilizing an advanced algorithm model, an inversion result is strictly verified, abnormal conditions are effectively processed, the construction safety of the ultra-deep foundation pit is guaranteed, the engineering quality is improved, and a reliable basis is provided for engineering decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering, and in particular to an algorithm model-based working condition inversion system and method under an ultra-deep foundation pit. Background Art

[0002] With the rapid development of urban construction, ultra-deep foundation pit projects are becoming more and more common in various large-scale construction projects, such as the construction of basements for high-rise and super-high-rise buildings, the excavation of urban underground rail transit stations, etc. The depth of ultra-deep foundation pits is usually more than 10 meters, and even reaches tens of meters in some special projects. The construction environment is complex, and there are often existing buildings and underground pipelines around them, which places extremely high requirements on construction safety and deformation control.

[0003] During the construction of ultra-deep foundation pits, it is crucial to accurately grasp the real-time working conditions of the foundation pits. However, traditional foundation pit monitoring and analysis methods have many limitations. In terms of data collection, in the early days, only simple tools such as levels and theodolites were used to measure the displacement and settlement of foundation pits. This was not only inefficient, but also difficult to achieve in terms of accuracy to meet the needs of modern ultra-deep foundation pit projects. When measuring groundwater levels, manual regular measurements cannot obtain real-time changes in water levels. There is a serious lag in monitoring key parameters such as the rate of change of water levels, making it difficult to issue early warnings when water levels change rapidly.

[0004] In the past, there was a lack of systematic processing procedures and quantitative evaluation indicators for geological data processing. Most engineers relied on their experience to conduct qualitative analysis of the mechanical parameters of the soil layers in the geological survey report, and were unable to accurately evaluate the comprehensive impact of geological conditions on the stability of the foundation pit. Faced with complex geological conditions, such as multiple layers of soil with different properties, areas with abundant groundwater and frequent water level changes, traditional methods are difficult to accurately judge the stability of the soil, resulting in frequent safety accidents caused by geological problems during construction.

[0005] There are also deficiencies in the analysis of construction monitoring data. In the past, the safety status of the foundation pit was mainly judged by comparing the monitoring data with the allowable values ​​in the specifications. This single evaluation method ignores the inherent connection between the data. There is a mutually influential and interrelated relationship between the displacement of the retaining structure and the supporting axial force, but the traditional method has failed to fully explore this relationship and cannot timely discover potential safety hazards. When the displacement of the retaining structure is abnormal, if it is judged only based on the displacement data, the impact of the change in the supporting axial force on the displacement may be ignored, thereby delaying the opportunity to take effective measures.

[0006] For the inversion of foundation pit working conditions, the traditional method based on empirical formulas and simple mechanical models is difficult to take into account the complex factors in the foundation pit construction process. When simulating the mechanical response of the foundation pit, it is impossible to accurately reflect the rheological properties of the soil, that is, the characteristics of the deformation of the soil under long-term loads over time. At the same time, the simulation of the step-by-step excavation effect of the foundation pit is not accurate enough, and it is impossible to perform dynamic analysis based on the actual conditions at different stages of the construction process, resulting in a large deviation between the inversion results and the actual working conditions, and unable to provide a reliable basis for construction decisions. Summary of the invention

[0007] The purpose of the present invention is to provide a system and method for inversion of working conditions under ultra-deep foundation pits based on an algorithm model, so as to solve the problems raised in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a working condition inversion system under an ultra-deep foundation pit based on an algorithm model, the system comprising: A data acquisition module is used to collect geological data and construction monitoring data of ultra-deep foundation pits, wherein the geological data includes soil layer mechanical parameters and groundwater level information, and the construction monitoring data includes retaining structure displacement and support axial force data; The geological data processing module is used to pre-process the collected geological data, compare the processed data with a preset first threshold value, generate a geological stability index from the processed geological data, if the geological stability index exceeds the first threshold value, the system determines that the geological conditions are unstable and suspends subsequent inversion processing; if the geological stability index is lower than the first threshold value or is within the range of the first threshold value, the geological conditions are determined to be stable, and the system continues to process the monitoring data; The monitoring data comparison module is used to analyze the collected displacement and support axial force data of the enclosure structure and generate a monitoring data evaluation index; compare the monitoring data evaluation index with the preset second threshold value to determine whether the current monitoring data is within the normal range; if the monitoring data evaluation index exceeds the second threshold value, the current working condition is judged to be abnormal, and the system suspends the model operation and issues a prompt; if the monitoring data evaluation index is lower than the second threshold value or is within the range of the second threshold value, the current working condition is judged to be normal, and the system continues data processing; The model operation module is used to input the processed geological data and monitoring data into the pre-built working condition inversion algorithm model for operation, and obtain the real-time working condition inversion results of the ultra-deep foundation pit; The inversion result verification module is used to verify the inversion results obtained by model operation and judge the accuracy of the inversion results by comparing with historical similar working condition data and theoretical calculation results; The exception handling module is used to handle abnormal situations during model operation and inversion result verification; if the system detects data anomalies or excessive deviations in inversion results during operation or verification, the exception handling module will suspend the current processing and re-evaluate the geological data and monitoring data.

[0009] Preferably, the soil layer mechanical parameters include soil elastic modulus and internal friction angle; the groundwater level information includes water level change rate; The soil stability index is proposed as The soil stability index is used to evaluate the soil's ability to resist deformation and damage and calculate the soil elastic modulus. , internal friction angle The comprehensive impact value and water level change rate Impact:

[0010] in It represents the soil stability index; To adjust the coefficient, control the influence of each parameter on the index; Indicates the elastic modulus of different soil layers; represents the internal friction angle of different soil layers; is the maximum value of the water level change rate, is the minimum value of the water level change rate, Reflects the magnitude of water level changes.

[0011] Preferably, by combining the soil stability index and groundwater level impact index Assess the comprehensive impact of geological conditions on ultra-deep foundation pits and obtain the geological stability index :

[0012] in It is the geological stability index, which reflects the comprehensive impact of geological conditions on ultra-deep foundation pits; is the adjustment coefficient, which controls the weight of the influence of soil stability and groundwater level on geological stability; It is a comprehensive index of groundwater level impact, which is calculated by the groundwater level depth. , water level change rate and the aquifer permeability The calculation results in:

[0013] is the adjustment coefficient, , , They are the groundwater level depth, water level change rate and aquifer permeability coefficient corresponding to different monitoring points respectively.

[0014] Preferably, the enclosure structure displacement data includes displacement change acceleration; the support axial force data includes axial force change frequency; The proposed displacement stability index is The displacement stability index is used to measure the stability of the displacement of the enclosure structure and calculate the acceleration of displacement change in each period. and displacement change ratio :

[0015] in represents the displacement stability index, It represents the adjustment coefficient, which controls the influence of the acceleration and displacement change amplitude on the index; Indicates the acceleration of displacement change at different time periods; represents the maximum displacement of the enclosure structure, represents the minimum displacement of the enclosure structure, It represents the displacement change amplitude ratio and evaluates the range of displacement change; The proposed positioning axial force stability index is The axial force stability index is used to evaluate the stability of the supporting axial force and calculate the frequency of axial force changes in each period. and axial force variation ratio :

[0016] in represents the axial force stability index; It represents the adjustment coefficient, which controls the influence of the frequency and amplitude of the axial force change on the index; Indicates the frequency of axial force changes in different periods; represents the maximum value of the supporting axial force, represents the minimum value of the supporting axial force, Indicates the axial force variation ratio and evaluates the range of axial force variation.

[0017] Preferably, the nonlinear combination displacement stability index , axial force stability index To evaluate the overall stability of the monitoring data, we can get the monitoring data evaluation index ; By combining the Euclidean distance and product of the two indexes, the comprehensive impact of various factors of the monitoring data can be reflected:

[0018] in It represents the monitoring data evaluation index, which is used to reflect the overall stability of the monitoring data; It represents the adjustment coefficient, which controls the comprehensive influence weight of displacement and axial force factors on the evaluation of monitoring data; represents the displacement stability index; represents the axial force stability index; It represents the Euclidean distance between the two, reflecting the joint deviation of various factors in the monitoring data; It represents the product of the two, amplifying the effect when both displacement and axial force are unstable at the same time.

[0019] Preferably, the working condition inversion algorithm model includes a deep learning model based on a spatiotemporal attention mechanism and an improved finite element model; The deep learning model based on the spatiotemporal attention mechanism is used to learn the spatiotemporal characteristics in the monitoring data, assign different weights to data at different times and locations, and explore the patterns in the data; the improved finite element model combines the rheological properties of the soil and the step-by-step excavation effect during the foundation pit construction process to simulate the mechanical response of the foundation pit.

[0020] Preferably, the deep learning model based on the spatiotemporal attention mechanism adopts a multi-head attention mechanism to process the displacement and axial force data in parallel, and the calculation formula is:

[0021] in, , , They are query, key, and value matrices respectively; is the number of attention heads , is the dimension of the key vector; is the output weight matrix.

[0022] Preferably, the improved finite element model introduces rheological terms and construction step influencing factors, and the improved finite element equation is:

[0023] in, is the stiffness matrix, is the displacement vector, is the relaxation function, reflecting the rheological characteristics of soil; is the external force vector, is the construction step influencing factor, For the The external force vector corresponding to the construction step, is the total number of construction steps.

[0024] Preferably, the inversion result verification module adopts a comparative verification method and a statistical test method; The comparative verification method is to compare the inversion results with the actual operating data under similar historical conditions and calculate the similarity between the two; the statistical test method is to perform hypothesis testing on the inversion results to determine whether the inversion results conform to a certain statistical distribution.

[0025] Preferably, a method for using the above-mentioned working condition inversion system comprises the following steps: The steps of the method of use include: The data acquisition module is used to collect geological data and construction monitoring data of ultra-deep foundation pits. The geological data includes soil mechanical parameters and groundwater level information, and the construction monitoring data includes retaining structure displacement and support axial force data. The geological data processing module is used to pre-process the collected geological data and calculate the soil stability index. and groundwater level impact index , and then the geological stability index is obtained , and compare it with the preset first threshold; if If the first threshold is exceeded, the geological conditions are judged to be unstable and the subsequent inversion processing is suspended; If it is lower than or within the first threshold, the geological conditions are determined to be stable and the next step is continued; Use the monitoring data comparison module to analyze the collected enclosure structure displacement and support axial force data, and calculate the displacement stability index respectively and axial force stability index , and then obtain the monitoring data evaluation index through nonlinear combination ,Will Compare with the preset second threshold; if If the second threshold is exceeded, the current operating condition is judged to be abnormal, the model operation is suspended and a prompt is issued; If it is lower than or within the second threshold value, the current working condition is determined to be normal and the next step is continued; The processed geological data and monitoring data are input into the pre-built working condition inversion algorithm model for calculation to obtain the real-time working condition inversion results of the ultra-deep foundation pit; Use the inversion result verification module to verify the inversion results obtained by model operation using comparative verification method and statistical test method; During the model operation and inversion result verification process, if data anomalies or excessive deviations in the inversion results are detected, the anomaly handling module will suspend the current processing and re-evaluate the geological data and monitoring data.

[0026] Compared with the prior art, the present invention has the following beneficial effects: The present invention can accurately obtain comprehensive and critical data of ultra-deep foundation pits through the data acquisition module, covering soil mechanical parameters, groundwater level information, retaining structure displacement, and support axial force data. The use of advanced sensor technology and scientific measurement methods greatly improves the accuracy and real-time performance of data acquisition, providing a reliable basis for subsequent analysis. For example, when monitoring the groundwater level, the water level change rate is recorded in real time. Compared with traditional manual regular measurements, it can capture small fluctuations in the water level in a timely manner and discover potential risks in advance. The geological data processing module pre-processes the collected geological data and comprehensively evaluates the geological conditions by calculating the soil stability index and the geological stability index. This quantitative analysis method overcomes the subjectivity and limitations of traditional empirical judgments and can accurately judge whether the geological conditions are stable. When encountering complex geological conditions, the construction plan can be adjusted in time according to the index changes to effectively prevent safety accidents caused by geological problems.

[0027] The monitoring data comparison module can evaluate the foundation pit construction conditions in real time and comprehensively by calculating the displacement stability index, axial force stability index and monitoring data evaluation index. This multi-dimensional evaluation method comprehensively considers the inherent connection between the displacement of the retaining structure and the support axial force data, avoiding the one-sidedness of the traditional single evaluation method. Once the monitoring data evaluation index exceeds the preset threshold, the system immediately determines that the current working condition is abnormal and issues a prompt, allowing construction personnel to quickly detect potential risks and take corresponding measures in a timely manner, such as strengthening support, adjusting excavation progress, etc., to effectively ensure construction safety and reduce the probability of accidents.

[0028] The working condition inversion algorithm model integrates a deep learning model based on the spatiotemporal attention mechanism and an improved finite element model. The deep learning model can learn the spatiotemporal characteristics in the monitoring data, explore the potential laws of the data, and assign reasonable weights to the data at different times and locations. When analyzing the changing trends of the displacement of the retaining structure and the supporting axial force over time, it can accurately identify abnormal change points and their influencing factors. The improved finite element model takes into account the rheological properties of the soil and the step-by-step excavation effect of the foundation pit, which is more in line with the actual construction situation. When simulating the mechanical response of the foundation pit, it can accurately predict the deformation and stress state of the foundation pit at different construction stages. Compared with the traditional model, it greatly improves the accuracy of the working condition inversion and provides a more reliable basis for construction decisions.

[0029] The inversion result verification module uses comparative verification method and statistical test method to strictly verify the inversion results obtained by model operation. The comparative verification method compares the inversion results with historical similar working condition data, and the statistical test method judges the rationality of the inversion results from a statistical perspective. Through the combination of these two methods, the accuracy of the inversion results can be effectively tested and possible deviations can be discovered in a timely manner. If the inversion results are found to be unreliable, the abnormality handling module will re-evaluate the geological data and monitoring data, adjust the model parameters or re-collect data to ensure that the final inversion results are reliable and provide strong support for engineering construction.

[0030] The exception handling module monitors data anomalies and inversion result deviations in real time during model calculation and inversion result verification. Once an anomaly is detected, the current processing is immediately suspended and the geological data and monitoring data are re-evaluated. This timely exception handling mechanism can avoid the impact of erroneous data on the inversion results, ensure the stable operation of the system, and ensure that the entire working condition inversion system continues to work reliably in a complex engineering environment, thus ensuring the safety of ultra-deep foundation pit construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a working principle diagram of the working condition inversion system of the present invention; Figure 2 The diagram is a step-by-step diagram for calculating soil mechanical parameters and soil stability index; Figure 3 Diagram of the steps in calculating the evaluation index for monitoring data. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] See also Figure 1-Figure 3 , the present invention provides a technical solution: a working condition inversion system under an ultra-deep foundation pit based on an algorithm model, which collects key data of the ultra-deep foundation pit through a data acquisition module. In terms of geological data acquisition, professional geological exploration equipment and technology are used to obtain soil layer mechanical parameters and groundwater level information. For example, drilling coring technology is used to obtain soil layer samples at different depths, and geotechnical tests are performed in the laboratory to determine parameters such as the elastic modulus and internal friction angle of the soil; a water level monitor is used to monitor groundwater level changes in real time, and record information such as the water level change rate. For construction monitoring data, displacement sensors and axial force sensors are used to monitor the displacement of the retaining structure and the support axial force in real time, respectively, to ensure the accuracy and timeliness of the data.

[0034] The collected data is transmitted to the geological data processing module for preprocessing. The preprocessing operation includes data cleaning to remove abnormal data caused by sensor failure, external interference and other factors; data standardization to make data of different magnitudes and types comparable. Afterwards, the processed data is used to generate a geological stability index according to a specific algorithm and compared with a pre-set first threshold. If the geological stability index exceeds the first threshold, it indicates that the geological conditions are unstable. At this time, the system suspends subsequent inversion processing to avoid unreliable inversion results due to abnormal geological conditions; if the index is at or below the first threshold, the geological conditions are determined to be stable, and the system continues the subsequent monitoring data processing process.

[0035] The monitoring data comparison module receives the displacement and support axial force data of the enclosure structure from the data acquisition module and conducts in-depth analysis. By calculating characteristic parameters such as displacement change acceleration and axial force change frequency, the displacement stability index and axial force stability index are generated in combination with the corresponding formula, and then the monitoring data evaluation index is obtained through nonlinear combination. The evaluation index is compared with the preset second threshold. If it exceeds the second threshold, the current working condition is judged to be abnormal, the system suspends the model operation and issues an alarm to remind relevant personnel; if the index is at or below the second threshold, the current working condition is judged to be normal, and the system continues to process data.

[0036] The model operation module inputs the data processed by the geological data processing module and the monitoring data comparison module into the pre-built working condition inversion algorithm model for operation. The algorithm model integrates the deep learning model based on the spatiotemporal attention mechanism and the improved finite element model. The deep learning model is responsible for mining the spatiotemporal characteristics in the monitoring data, giving reasonable weights to the data at different times and locations, so as to discover the potential laws behind the data; the improved finite element model combines the rheological properties of the soil and the step-by-step excavation effect during the foundation pit construction process to accurately simulate the mechanical response of the foundation pit, and finally obtain the real-time working condition inversion results of the ultra-deep foundation pit.

[0037] The inversion result verification module uses the comparative verification method and the statistical test method to verify the accuracy of the inversion results obtained by the model operation. The comparative verification method is to compare the inversion results with the actual working condition data under similar historical working conditions, calculate the similarity between the two, and evaluate the reliability of the inversion results; the statistical test method is to conduct hypothesis testing on the inversion results to determine whether they conform to a specific statistical distribution and verify the rationality of the inversion results from a statistical perspective.

[0038] During the entire system operation, the exception handling module monitors the model operation and inversion result verification process in real time. Once data anomalies (such as data missing, data mutation, etc.) or excessive deviations in inversion results are detected, the exception handling module will immediately suspend the current processing flow and re-evaluate the geological data and monitoring data. By re-checking the data acquisition equipment and applying data repair algorithms, the quality of the data is ensured, and the stable operation of the system and the accuracy of the inversion results are guaranteed.

[0039] The present invention will be further described below in conjunction with Examples 1 to 5: Example 1: In actual engineering, obtaining accurate soil mechanical parameters is the basis for calculating the soil stability index. Soil elastic modulus It reflects the ability of soil to resist elastic deformation and can be obtained in various ways. Common in-situ test methods include plate load test. In a suitable location is selected at the foundation pit site, a load-bearing plate of a certain size is placed on the surface of the soil layer, vertical loads are applied in stages, the settlement of the load-bearing plate is measured, and the elastic modulus of the soil is calculated according to the load-settlement curve, relevant specifications and empirical formulas. Assume that in a certain ultra-deep foundation pit project, three plate load tests were carried out on soil layers at different depths to obtain the elastic modulus of different soil layers. .

[0040] Internal friction angle It reflects the friction characteristics between soil particles and is generally measured through indoor triaxial shear tests. Original soil samples are collected from different soil layers, and the soil samples are made into standard specimens in the laboratory. They are placed in a triaxial shear tester, and different confining pressures and vertical pressures are applied to simulate the stress state of the soil in actual engineering. The stress-strain data when the soil sample is destroyed is recorded, and then the internal friction angle is calculated. For example, the internal friction angle of the soil samples corresponding to the three soil layers above is obtained through the test. .

[0041] Water level change rate in groundwater level information , by arranging multiple water level observation wells around the foundation pit for monitoring. The water level height needs to be measured regularly in the observation wells. Assume that the water level is measured every 2 hours during a certain monitoring period. In one of the observation wells, the initial water level height is After 2 hours, the water level becomes , then the water level change rate during this period is The maximum value of the water level change rate can be obtained during multiple observation periods. and minimum value , assuming that during a period of monitoring, . Soil stability index Used to evaluate the ability of soil to resist deformation and damage. The calculation formula is:

[0042] in, is the adjustment coefficient, and its value needs to be determined by comprehensive consideration of engineering geological conditions and experience. The value is usually 0.6; in hard soil, It can be adjusted to 0.4 appropriately. This embodiment takes soft soil as an example. Take 0.6, (Corresponding to the three soil layers mentioned above). The calculation process is as follows:

[0043]

[0044] Through the above detailed calculation process, the soil stability index is obtained, which can quantify the stability of the soil in the current state and provide key data support for subsequent geological stability assessment and engineering decision-making.

[0045] Example 2: Based on the results of Example 1, this example further combines the comprehensive index of groundwater level influence to calculate the geological stability index, and comprehensively and comprehensively evaluates the impact of geological conditions on ultra-deep foundation pits.

[0046] Calculation of a comprehensive index of groundwater level impact Need to get the depth of groundwater level , water level change rate and the aquifer permeability . Depth of groundwater level The water level observation well can be used for accurate measurement. Multiple observation points can be arranged at different locations around the foundation pit to obtain the groundwater level depth data at different locations. For example, at three observation points, the groundwater level depths are measured to be , , . Water level change rate The monitoring data in Example 1 are used. Aquifer Permeability Coefficient It can be determined through on-site pumping tests or reference to regional geological data. Assume that through the pumping test, the permeability coefficients of the aquifers corresponding to the three observation points are , , .

[0047] The calculation formula of the comprehensive index of groundwater level impact is:

[0048] in, is the adjustment coefficient, which is determined according to the hydrogeological conditions of the project site. The general value is 0.07; in water-poor formations, The value is 0.05. This example assumes that the formation is rich in water. Take 0.07, The calculation process is as follows:

[0049] Geological stability index By combining soil stability index and groundwater level impact index The calculation formula is:

[0050] in, is the adjustment coefficient, which is used to control the weight of the influence of soil stability and groundwater level on geological stability. In areas where geological conditions are greatly affected by soil stability, The value can be 1.1; in areas greatly affected by groundwater level, The value can be 0.9. This example assumes that the geological conditions in the region are relatively balanced by the influence of both. Take 1. and the calculated value in this embodiment Substituting into the formula, the calculation process is as follows:

[0051] Geological stability index The impact of two key factors, soil and groundwater level, on ultra-deep foundation pits is comprehensively considered. By comparing with the pre-set first threshold, it is possible to accurately judge whether the geological conditions are stable, providing a comprehensive and important basis for the safety assessment of engineering construction.

[0052] Example 3: In the actual construction scenario of the ultra-deep foundation pit, the monitoring of the displacement data of the retaining structure adopts the collaborative working mode of high-precision displacement sensors and acceleration sensors. The displacement sensors need to be reasonably arranged in the key parts of the retaining structure according to the design requirements of the foundation pit, such as at the corners of the foundation pit, the midpoints of the long sides and other locations that are prone to large displacements. These sensors monitor the displacement of the retaining structure in real time at set time intervals, such as 5 minutes. The acceleration sensor synchronously collects the acceleration data of the displacement change.

[0053] Assume that in a specific monitoring period, 5 sets of data are collected by the sensor. In the first 5-minute period, the displacement change acceleration for ; The second period, for The third period, for ; The fourth period, for ; The fifth period, for At the same time, during the entire monitoring process, the maximum displacement of the enclosure structure was recorded by the displacement sensor. 40mm, minimum is 10 mm, from which the displacement change ratio can be calculated as .

[0054] Proposed displacement stability index The calculation formula is:

[0055] Among them, the adjustment coefficient The value of needs to be determined based on the specific type of enclosure structure and rich engineering experience. For common underground continuous wall enclosure structures, The value is generally between 0.7 and 0.9. In this embodiment, it is assumed that an underground continuous wall enclosure structure is used. Combined with the actual project situation, Take 0.8, is the number of monitoring periods, where The specific calculation process is as follows:

[0056] To obtain the support axial force data, axial force sensors are installed at the key stress points of the support structure, and their accuracy must meet the engineering monitoring requirements. The axial force sensors also collect data at a certain time interval, and the collected data is processed by data analysis software to calculate the axial force change frequency.

[0057] In the same monitoring period, the axial force change frequencies in the five periods were calculated to be: 0.2Hz, 0.25Hz, 0.22Hz, is 0.26Hz, The maximum value of the support shaft force during the entire monitoring period is 0.23Hz. 600kN, minimum is 200kN, and the axial force variation ratio is .

[0058] The stability index of the proposed positioning axis force The calculation formula is:

[0059] Among them, the adjustment coefficient The value is determined based on the material properties of the supporting structure and engineering experience. For steel supporting structures, Usually in the range of 0.5-0.7; for concrete support structures, Generally, it is between 0.6 and 0.8. In this embodiment, it is assumed that a concrete support structure is used, and various factors are comprehensively considered. Take 0.7, The calculation process is as follows:

[0060] By accurately calculating the displacement stability index and axial force stability index , which can accurately quantify the stability of the retaining structure displacement and supporting axial force. These two indexes provide key data support for the subsequent evaluation of the overall stability of monitoring data, helping engineering personnel to promptly detect potential risks that may arise in the retaining structure and supporting system during the construction process, so as to take corresponding measures to ensure the safety of ultra-deep foundation pit construction.

[0061] Example 4: This example is mainly based on the displacement stability index obtained in Example 3 and axial force stability index , by calculating the monitoring data evaluation index through nonlinear combination , thereby making a comprehensive assessment of the overall stability of the monitoring data and providing a key basis for judging whether the current foundation pit construction conditions are normal.

[0062] It is known that the displacement stability index calculated in Example 3 is , axial force stability index Monitoring data evaluation index It is calculated by combining the Euclidean distance and product of the displacement stability index and the axial force stability index. The calculation formula is:

[0063] Among them, η is the adjustment coefficient, which is used to control the comprehensive influence weight of displacement and axial force factors on monitoring data evaluation. In foundation pit engineering, if the displacement of the retaining structure has a greater impact on the overall stability, η can be taken as 0.6; if the supporting axial force has a greater impact, η can be taken as 0.4. Assuming that the impact of the two is relatively balanced in this project, η is taken as 0.5.

[0064] First calculate and Values:

[0065] Then substitute the above results into the monitoring data evaluation index formula:

[0066] Monitoring Data Evaluation Index This reflects the overall stability of the monitoring data. Compare with the preset second threshold value. If If the second threshold is exceeded, it indicates that the current monitoring data is abnormal and there may be problems with the foundation pit construction conditions, and timely measures need to be taken to deal with them; If the value is at or below the second threshold, the current working condition is determined to be normal, and subsequent construction monitoring and data processing can continue. In this way, the ultra-deep foundation pit construction process can be effectively monitored in real time and risk warning can be carried out.

[0067] Embodiment 5: The working condition inversion algorithm model consists of a deep learning model based on the spatiotemporal attention mechanism and an improved finite element model. The deep learning model based on the spatiotemporal attention mechanism uses a multi-head attention mechanism to process the displacement and axial force data in parallel. The calculation formula is:

[0068] in, They are query, key, and value matrices respectively; is the number of attention heads; , is the dimension of the key vector; is the output weight matrix.

[0069] In any practical application, it is assumed that the collected displacement and axial force data are pre-processed to form time series data, which are divided according to a certain time step and feature dimension and constructed into a tensor form suitable for model input. Initial data of the matrix. Set the number of attention heads , the dimension of the key vector The deep learning model automatically assigns different weights to the data by learning the displacement and axial force data at different times and locations, and mines the spatiotemporal characteristics and potential laws in the data, such as discovering the changing trends and mutual relationships of the displacement and axial force of the foundation pit at different construction stages. The improved finite element model introduces rheological terms and construction step influencing factors, and the improved finite element equation is:

[0070] in, is the stiffness matrix, is the displacement vector, is the relaxation function (reflecting the rheological properties of soil), is the external force vector, is the construction step influencing factor, For the The external force vector corresponding to the construction step, is the total number of construction steps.

[0071] In the simulation of ultra-deep foundation pit engineering, the elastic modulus, Poisson's ratio and other parameters of the soil are determined according to the engineering geological survey report, and the geometric model and finite element mesh of the foundation pit are constructed. Combined with the actual construction plan of the foundation pit, the total number of construction steps is determined. , as well as the excavation depth of each construction step, support setting and other information, and then determine the construction step influencing factors and the external force vector The improved finite element equations were solved by numerical calculation methods to simulate the mechanical response of the foundation pit during the construction process, and the displacement, stress and other data of various parts of the foundation pit were obtained.

[0072] The inversion result verification module uses the comparative verification method and statistical test method to verify the inversion results obtained by model operation. The comparative verification method is to compare the inversion results with the actual working condition data under similar historical working conditions and calculate the similarity between the two. Assume that in similar historical working conditions, the actual displacement value of a specific position of the foundation pit retaining structure at a certain moment is The inversion displacement value obtained by the working condition inversion algorithm model of the present invention is , using Euclidean distance to calculate similarity :

[0073]

[0074] The higher the similarity, the closer the inversion result is to the actual situation and the higher the accuracy of the model.

[0075] The statistical test method is to perform hypothesis testing on the inversion results to determine whether the inversion results conform to a certain statistical distribution. For example, assuming that the support axial force data obtained by inversion obeys a normal distribution, by calculating the sample mean , sample standard deviation , construct the test statistic (in is the assumed total mean value, is the sample size). According to the pre-set significance level (such as , check the standard normal distribution table to determine the critical value. If the value is within the acceptance domain, the inversion result is considered to be consistent with the assumed statistical distribution, which further verifies the rationality of the inversion result; if If the value is outside the acceptance domain, the model or data needs to be re-evaluated and analyzed.

[0076] Through the above detailed construction, operation and strict verification of the working condition inversion algorithm model, the accuracy and reliability of the ultra-deep foundation pit working condition inversion can be effectively improved, providing strong support for engineering construction safety management.

[0077] The present invention also includes a method for using the working condition inversion system, the method comprising the following steps: The data acquisition module is used to collect geological data and construction monitoring data of ultra-deep foundation pits. The geological data includes soil mechanical parameters and groundwater level information, and the construction monitoring data includes retaining structure displacement and support axial force data. The geological data processing module is used to pre-process the collected geological data and calculate the soil stability index. and groundwater level impact index , and then the geological stability index is obtained , and compare it with the preset first threshold; if If the first threshold is exceeded, the geological conditions are judged to be unstable and the subsequent inversion processing is suspended; If it is lower than or within the first threshold, the geological conditions are determined to be stable and the next step is continued; Use the monitoring data comparison module to analyze the collected enclosure structure displacement and support axial force data, and calculate the displacement stability index respectively and axial force stability index , and then obtain the monitoring data evaluation index through nonlinear combination ,Will Compare with the preset second threshold; if If the second threshold is exceeded, the current operating condition is judged to be abnormal, the model operation is suspended and a prompt is issued; If it is lower than or within the second threshold value, the current working condition is determined to be normal and the next step is continued; The processed geological data and monitoring data are input into the pre-built working condition inversion algorithm model for calculation to obtain the real-time working condition inversion results of the ultra-deep foundation pit; Use the inversion result verification module to verify the inversion results obtained by model operation using comparative verification method and statistical test method; During the model operation and inversion result verification process, if data anomalies or excessive deviations in the inversion results are detected, the anomaly handling module will suspend the current processing and re-evaluate the geological data and monitoring data.

[0078] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0079] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A working condition inversion system under ultra-deep foundation pit based on algorithm model, characterized in that: The system comprises: A data acquisition module is used to collect geological data and construction monitoring data of ultra-deep foundation pits, wherein the geological data includes soil layer mechanical parameters and groundwater level information, and the construction monitoring data includes retaining structure displacement and support axial force data; The geological data processing module is used to pre-process the collected geological data, compare the processed data with a preset first threshold value, generate a geological stability index from the processed geological data, if the geological stability index exceeds the first threshold value, the system determines that the geological conditions are unstable and suspends subsequent inversion processing; if the geological stability index is lower than the first threshold value or is within the range of the first threshold value, the geological conditions are determined to be stable, and the system continues to process the monitoring data; The monitoring data comparison module is used to analyze the collected displacement and support axial force data of the enclosure structure and generate a monitoring data evaluation index; compare the monitoring data evaluation index with the preset second threshold value to determine whether the current monitoring data is within the normal range; if the monitoring data evaluation index exceeds the second threshold value, the current working condition is judged to be abnormal, and the system suspends the model operation and issues a prompt; if the monitoring data evaluation index is lower than the second threshold value or is within the range of the second threshold value, the current working condition is judged to be normal, and the system continues data processing; The model operation module is used to input the processed geological data and monitoring data into the pre-built working condition inversion algorithm model for operation, and obtain the real-time working condition inversion results of the ultra-deep foundation pit; The inversion result verification module is used to verify the inversion results obtained by model operation, and judge the accuracy of the inversion results by comparing with historical similar working condition data and theoretical calculation results; The exception handling module is used to handle abnormal situations during model operation and inversion result verification; if the system detects data anomalies or excessive deviations in inversion results during operation or verification, the exception handling module will suspend the current processing and re-evaluate the geological data and monitoring data.

2. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 1 is characterized in that: The soil layer mechanical parameters include soil elastic modulus and internal friction angle; the groundwater level information includes water level change rate; The soil stability index is proposed as The soil stability index is used to evaluate the soil's ability to resist deformation and damage and calculate the soil elastic modulus. , internal friction angle The comprehensive impact value and water level change rate Impact: ; in It represents the soil stability index; To adjust the coefficient, control the influence of each parameter on the index; Indicates the elastic modulus of different soil layers; represents the internal friction angle of different soil layers; is the maximum value of the water level change rate, is the minimum value of the water level change rate, Reflects the magnitude of water level changes.

3. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 2 is characterized in that: By combining soil stability index and groundwater level impact index Assess the comprehensive impact of geological conditions on ultra-deep foundation pits and obtain the geological stability index : ; in It is the geological stability index, which reflects the comprehensive impact of geological conditions on ultra-deep foundation pits; is the adjustment coefficient, which controls the weight of the influence of soil stability and groundwater level on geological stability; It is a comprehensive index of groundwater level impact, which is calculated by the groundwater level depth. , water level change rate and the aquifer permeability The calculations show that: ; is the adjustment coefficient, , , They are the groundwater level depth, water level change rate and aquifer permeability coefficient corresponding to different monitoring points respectively.

4. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 3 is characterized in that: The enclosure structure displacement data includes displacement change acceleration; the support axial force data includes axial force change frequency; The proposed displacement stability index is The displacement stability index is used to measure the stability of the displacement of the enclosure structure and calculate the acceleration of displacement change in each period. and displacement change ratio : ； in represents the displacement stability index, It represents the adjustment coefficient, which controls the influence of the acceleration and displacement change amplitude on the index; Indicates the acceleration of displacement change at different time periods; represents the maximum displacement of the enclosure structure, represents the minimum displacement of the enclosure structure, It represents the displacement change amplitude ratio and evaluates the range of displacement change; The proposed positioning axial force stability index is The axial force stability index is used to evaluate the stability of the supporting axial force and calculate the frequency of axial force changes in each period. and axial force variation ratio : ； in represents the axial force stability index; It represents the adjustment coefficient, which controls the influence of the frequency and amplitude of the axial force change on the index; Indicates the frequency of axial force changes in different periods; represents the maximum value of the supporting axial force, represents the minimum value of the supporting axial force, Indicates the axial force variation ratio and evaluates the range of axial force variation.

5. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 4 is characterized in that: By nonlinear combination of displacement stability index , axial force stability index To evaluate the overall stability of the monitoring data, we can get the monitoring data evaluation index ; By combining the Euclidean distance and product of the two indexes, the comprehensive impact of various factors of the monitoring data can be reflected: ； in It represents the monitoring data evaluation index, which is used to reflect the overall stability of the monitoring data; It represents the adjustment coefficient, which controls the comprehensive influence weight of displacement and axial force factors on the evaluation of monitoring data; represents the displacement stability index; represents the axial force stability index; It represents the Euclidean distance between the two, reflecting the joint deviation of various factors in the monitoring data; It represents the product of the two, amplifying the effect when both displacement and axial force are unstable at the same time.

6. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 5 is characterized in that: The working condition inversion algorithm model includes a deep learning model based on a spatiotemporal attention mechanism and an improved finite element model; The deep learning model based on the spatiotemporal attention mechanism is used to learn the spatiotemporal characteristics in the monitoring data, assign different weights to data at different times and locations, and explore the patterns in the data; the improved finite element model combines the rheological properties of the soil and the step-by-step excavation effect during the foundation pit construction process to simulate the mechanical response of the foundation pit.

7. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 6 is characterized in that: The deep learning model based on the spatiotemporal attention mechanism adopts a multi-head attention mechanism to process the displacement and axial force data in parallel. The calculation formula is: ; in, , , They are query, key, and value matrices respectively; is the number of attention heads , is the dimension of the key vector; is the output weight matrix.

8. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 7 is characterized in that: The improved finite element model introduces rheological terms and construction step influencing factors, and the improved finite element equation is: ； in, is the stiffness matrix, is the displacement vector, is the relaxation function, reflecting the rheological characteristics of soil; is the external force vector, is the construction step influencing factor, For the The external force vector corresponding to the construction step, is the total number of construction steps.

9. The working condition inversion system under the ultra-deep foundation pit based on the algorithm model according to claim 8 is characterized in that: The inversion result verification module adopts comparative verification method and statistical test method; The comparative verification method is to compare the inversion results with the actual working condition data under similar historical working conditions and calculate the similarity between the two; The statistical test method is to conduct hypothesis testing on the inversion results to determine whether the inversion results conform to a certain statistical distribution.

10. A method for using the working condition inversion system according to claim 9, characterized in that: The steps of the method of use include: The data acquisition module is used to collect geological data and construction monitoring data of ultra-deep foundation pits. The geological data includes soil mechanical parameters and groundwater level information, and the construction monitoring data includes retaining structure displacement and support axial force data. The geological data processing module is used to pre-process the collected geological data and calculate the soil stability index. and groundwater level impact index , and then the geological stability index is obtained , and compare it with the preset first threshold; if If the first threshold is exceeded, the geological conditions are judged to be unstable and the subsequent inversion processing is suspended; If it is lower than or within the first threshold, the geological conditions are determined to be stable and the next step is continued; Use the monitoring data comparison module to analyze the collected enclosure structure displacement and support axial force data, and calculate the displacement stability index respectively and axial force stability index , and then obtain the monitoring data evaluation index through nonlinear combination ,Will Compare with the preset second threshold; if If the second threshold is exceeded, the current operating condition is judged to be abnormal, the model operation is suspended and a prompt is issued; If it is lower than or within the second threshold value, the current working condition is determined to be normal and the next step is continued; The processed geological data and monitoring data are input into the pre-built working condition inversion algorithm model for calculation to obtain the real-time working condition inversion results of the ultra-deep foundation pit; Use the inversion result verification module to verify the inversion results obtained by model operation using comparative verification method and statistical test method; During the model operation and inversion result verification process, if data anomalies or excessive deviations in the inversion results are detected, the anomaly handling module will suspend the current processing and re-evaluate the geological data and monitoring data.

Images