A method and device for intelligent prediction of pile foundation bearing capacity

Abstract

This invention provides an intelligent prediction method and device for pile foundation bearing capacity. The method includes: forming in-situ test samples based on in-situ engineering geological tests; obtaining optimal geological parameters for karst geology through surrogate model inversion; acquiring geological parameters from field tests under multiple working conditions to form a pile foundation field test sample library; acquiring pile foundation bearing performance parameters for various karst geologies under scaled-down pile foundation model experiments to form a scaled-down pile foundation model sample library; performing simulations using the optimal geological parameters for each karst geology to form a pile foundation simulation sample library; constructing a pile foundation bearing capacity training sample library by combining the three sample libraries; training the pile foundation bearing capacity prediction model; and inputting construction site geological parameters into the pile foundation bearing capacity prediction model to optimize and adjust the pile foundation model at the actual construction site based on the obtained geological parameters. This invention integrates data from four research paradigms to construct a more comprehensive and accurate intelligent prediction method for pile foundation bearing capacity.

Description

Technical Field

[0001] This invention relates to the field of pile foundation bearing capacity monitoring technology, and in particular to an intelligent prediction method and device for pile foundation bearing capacity. Background Technology

[0002] In pile foundation tests in karst areas, scientific research typically employs the following main paradigms: model testing, simulation analysis, in-situ testing, and data-driven approaches.

[0003] Model testing involves creating a scaled-down karst model in a laboratory or specialized testing environment to simulate actual karst geological conditions. By observing and measuring the model's loading, stress, and deformation, a preliminary understanding of the pile foundation's behavior in karst regions can be obtained. This method can effectively reveal the pile foundation's response under specific geological conditions, but the results need to be verified through actual field testing.

[0004] Simulation analysis utilizes computer modeling and numerical simulation techniques to virtually test pile foundations in karst areas. This method simulates the behavior of pile foundations under different loads and geological conditions by establishing complex geological and structural models. By adjusting model parameters, the performance of pile foundations in various scenarios can be studied, thus providing important references for practical engineering design.

[0005] In-situ testing involves direct testing conducted in actual karst areas, such as static cone penetration tests (CPT) and in-situ shear tests. These tests provide accurate information about the physical and mechanical properties of the soil and rock, helping to assess the performance of pile foundations under real-world conditions. In-situ testing is generally highly realistic and accurate, but it is also more costly and complex.

[0006] Data-driven approaches involve analyzing large amounts of field-collected data, such as geological survey data and pile foundation load test data, combined with machine learning and data mining techniques, to predict and optimize the design and construction of pile foundations. This method can utilize historical and real-time data for model training, thereby improving the accuracy of predicting pile foundation behavior in karst areas and the scientific rigor of engineering decisions.

[0007] Although model experiments, simulation analysis, and in-situ testing generate a large amount of data, the research methods are relatively independent and have limited integration. Summary of the Invention

[0008] This invention provides an intelligent prediction method and device for pile foundation bearing capacity. By integrating data from four research paradigms, a more comprehensive and accurate intelligent prediction method for pile foundation bearing capacity is constructed.

[0009] According to one aspect of the present invention, an intelligent prediction method for pile foundation bearing capacity is provided, comprising:

[0010] Based on in-situ engineering tests, geological parameters of various karst geological types are obtained, and an in-situ test sample library is formed by using one of the geological parameters of the karst geological type as a sample.

[0011] Using the actual observation data of the karst geology, the optimal geological parameters of the karst geology are obtained by inversion based on the trained first pile foundation bearing capacity surrogate model;

[0012] Geological parameters from field tests under multiple working conditions were obtained to form a sample library of field tests for pile foundations;

[0013] The bearing capacity parameters of various karst geological conditions under the pile foundation scale model test were obtained to form a pile foundation scale model sample library;

[0014] Using the optimal geological parameters of various karst geological formations, simulations are performed on each of the karst geological formations to calculate the pile foundation bearing capacity parameters of each karst geological formation, thus forming a pile foundation simulation sample library.

[0015] The pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library are combined to construct a pile foundation bearing capacity training sample library;

[0016] The pile foundation bearing capacity prediction model was trained using the aforementioned pile foundation bearing capacity training sample library;

[0017] The optimal geological parameters are obtained by inputting the actual geological parameters from the construction site into the trained pile foundation bearing capacity prediction model.

[0018] The pile foundation model at the actual construction site was optimized and adjusted based on the optimal geological parameters.

[0019] Optionally, it also includes: comparing the geological parameters obtained from the in-situ civil engineering test with the test results of the scaled-down pile foundation model test; adjusting the method of the scaled-down pile foundation model test or adjusting the design of the pile foundation model according to the comparison results.

[0020] Optionally, it also includes: inverting the geological parameters of the various karst geological formations under reduced dimensions into the second pile foundation bearing capacity proxy model to obtain the optimal geological parameters of the various karst geological formations under reduced dimensions.

[0021] Optionally, before inverting the geological parameters of the various karst geological formations under reduced dimensions into the second pile foundation bearing capacity surrogate model to obtain the optimal geological parameters of the various karst geological formations under reduced dimensions, the following steps are also included:

[0022] Select training, test, and validation sets from the pile foundation scaled model sample library;

[0023] The second pile foundation bearing capacity surrogate model was trained using the training set;

[0024] The performance of the second pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the second pile foundation bearing capacity surrogate model is evaluated using the validation data.

[0025] The final performance of the second pile foundation bearing capacity proxy model is evaluated using the test set.

[0026] Optionally, it further includes: using the optimal geological parameters of various karst geological formations under scaled-down conditions to simulate each karst geological formation, calculating the pile foundation bearing capacity of each karst geological formation, obtaining a scaled-down simulation sample library, and supplementing the pile foundation scaled-down model sample library with the data from the scaled-down simulation sample library.

[0027] Optionally, it also includes: correcting the pile foundation simulation sample library obtained by simulation simulation based on the conversion relationship between the data of field tests on pile foundation bearing capacity of various karst geological conditions and the simulation data of each karst geological condition under the same working conditions.

[0028] Optionally, before obtaining the optimal geological parameters of the karst geology by inverting the data based on the trained first pile foundation bearing capacity surrogate model using the actual observation data of the karst geology, the process further includes:

[0029] The training set, test set, and validation set are selected from the in-situ test sample library;

[0030] The first pile foundation bearing capacity proxy model was trained using the training set;

[0031] The performance of the first pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the first pile foundation bearing capacity surrogate model is evaluated using the validation data.

[0032] The final performance of the first pile foundation bearing capacity proxy model is evaluated using the test set.

[0033] Optionally, before simulating each of the karst geological formations using the optimal geological parameters of various karst geological formations, calculating the pile foundation bearing capacity parameters of each karst geological formation, and forming a pile foundation simulation sample library, the method further includes:

[0034] The parameters of the third pile foundation bearing capacity surrogate model are adjusted by using the pile foundation field test sample library.

[0035] The optimal geological parameters are obtained by inversion using the third pile foundation bearing capacity proxy model.

[0036] Optionally, before training the pile foundation bearing capacity prediction model using a training sample library, the method further includes:

[0037] Data preprocessing, noise reduction, and feature extraction are performed on the data in the training sample library.

[0038] According to another aspect of the present invention, an intelligent prediction device for pile foundation bearing capacity is provided, comprising:

[0039] The first sample acquisition unit is used to acquire geological parameters of various karst geology based on in-situ engineering tests, and to form an in-situ test sample library by using one of the geological parameters of the karst geology as a sample.

[0040] The inversion unit is used to invert the karst geology based on the trained first pile foundation bearing capacity surrogate model using the actual observation data of the karst geology, so as to obtain the optimal geological parameters of the karst geology.

[0041] The second sample acquisition unit is used to acquire geological parameters of field tests under multiple working conditions to form a sample library of pile foundation field tests.

[0042] The third sample acquisition unit is used to acquire pile bearing capacity parameters of various karst geological conditions under the pile foundation scale model experiment, and form a pile foundation scale model sample library.

[0043] The fourth sample acquisition unit is used to simulate each of the karst geologies using the optimal geological parameters of various karst geologies, calculate the pile foundation bearing capacity parameters of each of the karst geologies, and form a pile foundation simulation sample library.

[0044] The fourth sample acquisition unit is used to construct a pile foundation bearing capacity training sample library by combining the pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library.

[0045] The training unit is used to train the pile foundation bearing capacity prediction model using the pile foundation bearing capacity training sample library.

[0046] The prediction unit is used to input the geological parameters of the pile foundation model at the actual construction site into the trained pile foundation bearing capacity prediction model to obtain the optimal geological parameters;

[0047] The model optimization unit is used to optimize and adjust the pile foundation model at the actual construction site based on the optimal geological parameters.

[0048] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0049] At least one processor; and

[0050] A memory communicatively connected to the at least one processor; wherein,

[0051] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to execute the intelligent prediction method for pile foundation bearing capacity according to any embodiment of the present invention.

[0052] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions, the computer instructions being configured to cause a processor to execute and implement the intelligent prediction method for pile foundation bearing capacity according to any embodiment of the present invention.

[0053] The technical solution of this invention integrates geological parameters from field tests under multiple working conditions to form a pile foundation bearing capacity training sample library. This library comprises a field test sample library, a scaled-down model sample library obtained from pile foundation scaled-down model experiments, and a simulation sample library obtained from simulation experiments. The pile foundation bearing capacity prediction model is then trained using these samples. Finally, the geological parameters of actual construction site pile foundation models are input into the trained model to obtain optimal geological parameters. Based on these optimal parameters, the actual construction site pile foundation models are further optimized and adjusted. This invention integrates data from four research paradigms to construct a more comprehensive and accurate intelligent prediction method for pile foundation bearing capacity.

[0054] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A flowchart of an intelligent prediction method for pile foundation bearing capacity provided in an embodiment of the present invention;

[0057] Figure 2 A flowchart illustrating another intelligent prediction method for pile foundation bearing capacity provided in an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the model test apparatus and rock strata distribution provided in one embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram of the arrangement of fiber optic cables and strain gauge sensors in a pile body according to one embodiment of the present invention;

[0060] Figure 5 This is a schematic diagram of the fiber optic cable layout on the top and side of a cave provided in one embodiment of the present invention;

[0061] Figure 6 This is a schematic diagram of the overall experimental apparatus in one embodiment of the present invention;

[0062] Figure 7 This is a schematic diagram of the structure of an intelligent prediction device for pile foundation bearing capacity provided in an embodiment of the present invention;

[0063] Figure 8 This is a schematic diagram of the structure of an electronic device for implementing an intelligent prediction method for pile foundation bearing capacity according to an embodiment of the present invention. Detailed Implementation

[0064] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0065] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0066] Figure 1 A flowchart of an intelligent prediction method for pile foundation bearing capacity is provided as an embodiment of the present invention. For example... Figure 1 As shown, the method includes:

[0067] S101. Based on in-situ engineering tests, obtain geological parameters of various karst geological types, and use one of the geological parameters of the karst geological type as a sample to form an in-situ test sample library.

[0068] Collect in-situ tests to obtain geological parameters of various karst geological formations. These geological parameters include the physical properties of the soil (such as density, water content, particle size distribution, etc.), mechanical properties (such as shear strength, compression modulus, etc.), structural characteristics of the rocks, and environmental data (such as groundwater level, ground temperature, etc.). A data sample can be formed by combining the geological parameters of a karst geological formation with the geological type of that karst geological formation, thus creating an in-situ test sample library containing the characteristics of the geological parameters and the target output.

[0069] S102. Using the actual observation data of the karst geology, the optimal geological parameters of the karst geology are obtained by inversion based on the trained first pile foundation bearing capacity proxy model.

[0070] The surrogate model for the bearing capacity of the first pile foundation is a predictive model constructed using a neural network model. It can be trained using data from an in-situ test sample library to build a surrogate model related to geological parameters. The purpose of the surrogate model is to predict relevant geological parameters based on the input test data.

[0071] Inversion analysis refers to using a surrogate model of the first pile foundation bearing capacity to infer geological parameters. In other words, it involves using the surrogate model to deduce actual geological parameters, which are then used to adjust parameters within the surrogate model to match actual observation data. Inversion analysis can identify which geological parameters best represent the characteristics of this type of karst geology.

[0072] Optimal geological parameters are typically those parameters that accurately describe the characteristics of karst geology in geological models and exhibit the best performance in practical applications. Specifically, optimal parameters minimize the error between the model's predictions and actual observations. Optimal geological parameters maintain the model's stability and reliability under different conditions. Furthermore, optimal parameters should accurately reflect the actual characteristics and behavior of karst geological bodies.

[0073] S103. Obtain geological parameters from field tests under multiple working conditions to form a sample library for pile foundation field tests.

[0074] Typical geological lengths can be selected for field pile foundation tests. This allows for the determination of parameters corresponding to geological conditions, pile types, and load requirements under different working conditions, including the characteristics of the karst environment, pile foundation design parameters, and the type and magnitude of the load. Representative test sites can be selected based on the corresponding geological parameters to complete field tests under those geological parameters and working conditions. During the field tests, the arrangement of test piles, load application methods, and data acquisition methods can be adjusted, and then the field tests are implemented according to the design scheme. Field tests can include static load tests or dynamic load tests, during which key data such as pile foundation settlement and applied loads need to be recorded. In the data analysis phase, the test data need to be analyzed in detail to evaluate the bearing capacity of the pile foundation under different working conditions and to verify the pile foundation bearing capacity model. If problems are found during the experiments, the design or construction process can be adjusted to improve the applicability and reliability of the pile foundation. Finally, the data of the corresponding geological parameters for each working condition are compiled as a sample to construct a field pile foundation test sample library.

[0075] S104. Obtain the pile bearing capacity parameters of various karst geological conditions under the pile foundation scale-down model test, and form a pile foundation scale-down model sample library.

[0076] Scaled-down model tests of pile foundations were conducted to determine their bearing capacity under different karst geological conditions. For example, the specific effects of factors such as the height and span of karst caves, the number of piles, pile length, and pile diameter on the bearing capacity of the pile foundation were investigated. By setting different geological parameters, the bearing capacity of the pile foundation under different geological parameters could be obtained.

[0077] Physically scaled model tests can include analyzing the relationship between pile top load and the failure of the rock strata at the top of the karst cave, as well as the decrease in adhesion between the pile foundation and the rock strata. Through physically scaled model tests combined with numerical analysis techniques, changes in the pile foundation bearing characteristics can be obtained, including the relationship between pile top load and the failure of the rock strata at the top of the karst cave, and the attenuation of adhesion between the pile foundation and the rock strata, thus forming a scaled model sample library. Data on the pile foundation bearing capacity of a karst geological formation under a scaled model are used as a sample, and the sample data are converted to construct a sample library of pile foundation bearing capacity from the model tests.

[0078] S105. Using the optimal geological parameters of various karst geological formations, simulate each karst geological formation separately, calculate the pile foundation bearing capacity parameters of each karst geological formation, and form a pile foundation simulation sample library.

[0079] Given the limitations of indoor model testing in terms of time and cost, conducting indoor model tests covering multiple working conditions is impractical. Therefore, this embodiment adopts a simulation method, relying on finite element modeling technology to simulate factors such as geological parameters, environmental conditions, and test conditions. By simulating various geological parameters and karst conditions, the bearing capacity of karst pile foundations is calculated, thereby constructing a simulation sample library.

[0080] S106. The pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library are combined to construct a pile foundation bearing capacity training sample library.

[0081] S107. The pile foundation bearing capacity prediction model is trained using the pile foundation bearing capacity training sample library.

[0082] S108. Input the actual geological parameters from the construction site into the trained pile foundation bearing capacity prediction model to obtain the optimal geological parameters.

[0083] This embodiment constructs a training sample library by combining the pile foundation field test sample library, the pile foundation scaled model sample library, and the pile foundation simulation sample library. This training sample library is used to train the pile foundation bearing capacity prediction model. The pile foundation bearing capacity prediction model can include BP neural network models, Kriging models, multinomial models, etc.

[0084] In one embodiment, before training the pile foundation bearing capacity prediction model, the data in the training sample library can be preprocessed, denoised, and feature extracted to ensure data consistency and accuracy, and to handle missing and outlier values ​​to accurately capture the strain of the pile foundation. Furthermore, the data in the training sample library can be dimensionally converted and scaled to ensure that the pile foundation bearing capacity prediction model can accurately learn the complex relationships between various parameters. After the pile foundation bearing capacity prediction model is trained, the calculation formula for the pile foundation bearing capacity can be further optimized based on the optimal geological parameters output.

[0085] The results output by the pile foundation bearing capacity prediction model are based not only on the comprehensiveness of experimental and analytical data, but also on the combined influence of various geological parameters on pile foundation behavior. A well-trained pile foundation bearing capacity prediction model can derive an accurate bearing capacity calculation formula applicable to pile foundations in karst areas. This formula comprehensively considers the actual situation of field test data and the theoretical prediction of finite element analysis.

[0086] In constructing the pile foundation bearing capacity prediction model, this embodiment employs supervised learning or statistical analysis methods to correlate the collected strain data with historical and measured pile foundation bearing capacity data. The characteristics of the strain data may include the amplitude, frequency, and rate of change of the strain signal. Effective selection of these features, such as linear regression, decision trees, support vector machines (SVM), and neural networks, is used to train and construct the pile foundation bearing capacity prediction model, thereby achieving accurate prediction of pile foundation bearing capacity. During the training phase of the pile foundation bearing capacity prediction model, cross-validation and parameter optimization techniques can also be used to improve the model's prediction accuracy and stability.

[0087] Furthermore, this embodiment allows for the validation of the pile foundation bearing capacity prediction model to evaluate its performance in predicting pile foundation bearing capacity. The validation process may employ techniques such as retain-set evaluation and cross-validation to ensure the model has good generalization ability and can effectively handle unseen data. Once the pile foundation bearing capacity prediction model passes validation, it can be applied to actual strain monitoring data to achieve real-time prediction and monitoring of pile foundation bearing capacity. Real-time collected strain data can be input into the trained pile foundation bearing capacity prediction model to evaluate the bearing capacity state of the pile foundation in real time.

[0088] S108. Optimize and adjust the pile foundation model at the actual construction site based on the optimal geological parameters.

[0089] During on-site construction, a detailed geological survey is first conducted to collect geological parameters and pile foundation bearing capacity data for the karst area. The geological parameters are then input into the pile foundation bearing capacity prediction model. Based on the prediction results of the model, a preliminary design of the pile foundation is carried out. The prediction results represent the bearing capacity of a single pile. The preliminary design requires designing a pile group consisting of multiple piles based on the bearing capacity of a single pile. The design of the pile group is then analyzed and optimized in detail. Finally, the design scheme of the pile foundation model is applied to on-site construction, and real-time monitoring and feedback are conducted to continuously optimize the model and design scheme.

[0090] The technical solution of this invention integrates geological parameters from field tests under multiple working conditions to form a pile foundation bearing capacity training sample library. This library comprises a field test sample library, a scaled-down model sample library obtained from pile foundation scaled-down model experiments, and a simulation sample library obtained from simulation experiments. The pile foundation bearing capacity prediction model is then trained using these samples. Finally, the geological parameters of actual construction site pile foundation models are input into the trained model to obtain optimal geological parameters. Based on these optimal parameters, the actual construction site pile foundation models are further optimized and adjusted. This invention integrates data from four research paradigms to construct a more comprehensive and accurate intelligent prediction method for pile foundation bearing capacity.

[0091] In one embodiment, the method further includes comparing the geological parameters obtained from the in-situ civil engineering test with the test results from the scaled-down pile foundation model test; and adjusting the method of the scaled-down pile foundation model test or adjusting the design of the pile foundation model based on the comparison results.

[0092] Specifically, scaled-down model tests for pile foundations can use materials similar to those used in in-situ civil engineering tests. When selecting materials, their mechanical properties and other characteristics relevant to the research need to be considered. In this embodiment, a set of model tests was conducted based on research on in-situ tests in karst areas, with the ratio of the model test to the in-situ test set at 1:30. In the model test, materials similar to those used in in-situ civil engineering tests were selected; for example, a roofless, thickened, reinforced plastic water tank could be chosen, reinforced with a stainless steel frame on the outside. The karst cave was buried at a depth of 50cm, and its location was 15cm. 3 The cubic blocks were filled, representing unfilled karst caves. A three-pile cap foundation was selected, with 4cm diameter acrylic columns used to simulate the piles, embedded 8cm into the rock strata. The geological conditions were simplified, with different materials used for pouring at different depths. The mix proportions of similar materials were determined based on preliminary tests. Moderately weathered marl, at 47cm, was located in the lower layer of the model box to simulate the bedrock and embedded rock sections of the prototype; strongly weathered marl, at 30cm, simulated the overlying soil layer of the prototype. A schematic diagram of the rock strata depth and model is shown below. Figure 3 As shown, Figure 3 (a) is a schematic diagram of the model test device of the present invention. Figure 3 (b) is a schematic diagram of the rock strata distribution. The model materials selected for the experiment were determined based on reference data provided by previous similar material experiments on marl. River sand was used as aggregate, cement / gypsum as the main cementing material, diatomaceous earth / red clay / marl powder as conditioning material, and the amount of mixing water was 1 / 7 of the mixing material. The similarity ratio of the simulated material density was 1.5, the similarity ratio of the internal friction angle was 1, and the similarity ratio of the strain was 1.

[0093] In model tests, sensors and instruments are installed to measure the required physical quantities. In this case, fiber optic sensors and strain gauge sensors are deployed on the pile body, and fiber optic sensors are also deployed on the sides and top of the karst cave. To obtain pile strain data, in addition to using traditional resistance strain gauges, fiber optic sensors are also deployed on the pile body. To avoid data loss due to strain gauge deactivation during pouring and loading, strain gauges are symmetrically arranged at the same elevation to ensure data integrity. Details of the pile body resistance strain gauge deployment are as follows... Figure 4 As shown, Figure 4 (a) and (b) are schematic diagrams of the three-dimensional structure of the pile. Figure 4(c) is a schematic diagram of the strain gauge sensors deployed on the pile. Before model casting, strain gauge sensors were deployed on the pile according to the test plan. Four layers were laid within the thickness of the top slab of the karst cave, with 50cm long optical fibers connecting each layer. The optical fibers on the sides of the karst cave were fixed to a non-strength fiber mesh according to the deployment plan, facilitating laying and positioning during model casting. The optical fiber deployment is as follows: Figure 5 As shown, Figure 5 (a)-(c) show schematic diagrams of the fiber optic cable layout on the top and sides of the karst cave. In the model test, a static strain acquisition instrument and an AQ8603 fiber optic strain analyzer were used for data acquisition. After the model was poured, it was kept at a certain humidity and statically cured for 21 days before subsequent loading tests. During the model curing period, data was collected daily from the pile strain sensors and the fiber optic sensors on the top and sides of the karst cave to determine the sensor survival status. The model test loading system mainly consists of a reaction frame, force transmission column, hydraulic jacks, and a hydraulic pump; the measurement system consists of a digital dial indicator and a load sensor with a display screen. Pre-loading was performed on the pile to ensure close contact between the pile and the jacks, and initial loading data was collected and recorded. Initial loading was performed in stages, and later, based on the actual test conditions and load-displacement curves, staged loading was performed again. Data was collected during the load holding period until the model was destroyed. The integrity of the fiber optic monitoring range was ensured, and the necessary data was recorded and monitored in real time. This data will be used to analyze and compare the results of the model test with the results of the in-situ test. The results of the model test were analyzed and compared with the results of the in-situ test. Assess the consistency and differences between the two. Based on the comparison results, the model design or experimental methods can be adjusted to improve the model's accuracy and reliability.

[0094] Collect and organize the data on pile top displacement, strain, and karst cave strain from in-situ and model tests again. Process the data using appropriate engineering or geological analysis tools, such as determining load-displacement curves and stress-strain curves. Hydraulic jacks are recommended for loading the test. The loading and unloading methods should comply with the following regulations: 1. Loading should be carried out in stages, with each stage using equal loading increments. The stage load should preferably be 1 / 10 of the maximum load value or the estimated ultimate bearing capacity, with the first stage load being twice the stage load; 2. Unloading should be carried out in stages, with each unloading increment preferably twice the stage load during loading, and unloading should be equal in increments; 3. During loading and unloading, the load transfer should be uniform, continuous, and without impact, and the variation in load level during maintenance should not exceed ±10% of the stage load. The test steps for the rapid maintenance load method are as follows: 1. After each load level is applied, maintain it for 1 hour, and measure the pile top settlement at 5 minutes, 15 minutes, and 30 minutes, and then measure it every 15 minutes thereafter. 2. When the cumulative measurement time is 1 hour, if the settlement increment of the pile top in the last 15-minute time interval does not converge significantly with the settlement increment of the pile top in the adjacent 15-minute time interval, the load maintenance time should be extended until the settlement increment in the last 15 minutes is less than the settlement increment in the adjacent 15 minutes. 3. The conditions for terminating the loading are: (1) Under a certain load, the settlement at the top of the pile is more than 5 times the settlement under the previous load, and the total settlement at the top of the pile exceeds 40mm; (2) Under a certain load, the settlement at the top of the pile is more than 2 times the settlement under the previous load, and the relative stability standard has not been reached after 24 hours; (3) The maximum load value required by the design has been reached and the settlement at the top of the pile has reached the relative stability standard; (4) When the engineering pile is used as an anchor pile, the pull-out of the anchor pile has reached the allowable value; (5) When the load-settlement curve is gradually changing, the load can be applied up to a total settlement at the top of the pile of 60mm to 80mm; when the pile end resistance has not been fully utilized, the load can be applied up to a cumulative settlement at the top of the pile of 80mm. 4. When unloading, each load level is maintained for 15 minutes. After measuring the settlement at the top of the pile at the 5th minute and 15th minute, the first load level can be unloaded. After unloading to zero, the residual settlement at the top of the pile should be measured for 1 hour, with measurements taken at 5 minutes, 15 minutes, and 30 minutes. When using the rapid method, different regions should summarize and accumulate experience, and may propose appropriate settlement relative stability control standards based on local conditions.The ultimate bearing capacity of pile foundations can be determined based on the load-displacement curve. When determining the vertical compressive bearing capacity of a single pile, a load-settlement (QS) curve and a settlement-time logarithm (S-lgt) curve should be plotted. 1. Based on the characteristics of settlement variation with load: For a steeply sloping QS curve, the load value corresponding to the starting point of a significant steep drop should be taken. 2. Based on the characteristics of settlement variation with time: The load value of the previous stage before the obvious downward bend at the tail of the S-lgt curve should be taken. 3. For a gradually changing QS curve, the load value corresponding to S equal to 40mm should be taken based on the total settlement at the pile top; for piles with D (D is the pile tip diameter) greater than or equal to 800mm, the load value corresponding to S equal to 0.05D can be taken; when the pile length is greater than 40m, elastic compression of the pile body should be considered. Evaluate the reliability and applicability of the experimental data, such as determining parameters like material strength, stiffness, and plastic behavior. A specific structural diagram of the model test apparatus is shown below. Figure 6 As shown.

[0095] This embodiment also provides a flowchart of another intelligent prediction method for pile foundation bearing capacity, such as... Figure 2 As shown, Figure 2 The process also includes: inverting the geological parameters of various karst geological formations under reduced dimensions into the second pile foundation bearing capacity proxy model to obtain the optimal geological parameters of various karst geological formations under reduced dimensions.

[0096] The surrogate model for the bearing capacity of the second pile foundation is a predictive model constructed using a neural network model. It can be trained using data from a scaled model sample library to build a surrogate model related to geological parameters. The purpose of the surrogate model is to predict relevant geological parameters based on the input experimental data.

[0097] In one embodiment, before inverting the geological parameters of various karst geological formations at a reduced scale into the second pile foundation bearing capacity surrogate model to obtain the optimal geological parameters of the various karst geological formations at a reduced scale, the method further includes:

[0098] Select training, test, and validation sets from the pile foundation scaled model sample library;

[0099] The second pile foundation bearing capacity surrogate model was trained using the training set;

[0100] The performance of the second pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the second pile foundation bearing capacity surrogate model is evaluated using the validation data.

[0101] The final performance of the second pile foundation bearing capacity proxy model is evaluated using the test set.

[0102] Specifically, data cleaning, standardization, and feature selection can be performed on each sample in the scaled model sample library dataset. Training, testing, and validation data are selected from the scaled model sample library, and an appropriate neural network data model type is chosen as the first surrogate model for pile foundation bearing capacity. The model is then trained using the training data. Subsequently, the model performance is optimized by adjusting hyperparameters, and the model effect is evaluated using validation data, with further parameter adjustments as necessary. Finally, the final performance of the model is evaluated on an independent test set.

[0103] Figure 2 The embodiment of the intelligent prediction method for pile foundation bearing capacity shown further includes: using the optimal geological parameters of various karst geological formations under scaled-down conditions to simulate each karst geological formation, calculating the pile foundation bearing capacity of each karst geological formation, obtaining a scaled-down simulation sample library, and supplementing the pile foundation scaled-down model sample library with the data from the scaled-down simulation sample library.

[0104] The optimal geological parameters of various karst geological formations obtained under scaled-down conditions can be used to simulate and calculate the pile foundation bearing capacity, thereby obtaining a scaled-down simulation sample library to supplement the scaled-down model sample library.

[0105] Figure 2 In the embodiment of the intelligent prediction method for pile foundation bearing capacity shown, before the step of inverting the optimal geological parameters of the karst geology based on the trained first pile foundation bearing capacity surrogate model using the actual observation data of the karst geology, the method further includes:

[0106] The training set, test set, and validation set are selected from the in-situ test sample library;

[0107] The first pile foundation bearing capacity proxy model was trained using the training set;

[0108] The performance of the first pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the first pile foundation bearing capacity surrogate model is evaluated using the validation data.

[0109] The final performance of the first pile foundation bearing capacity proxy model is evaluated using the test set.

[0110] Specifically, data cleaning, standardization, and feature selection can be performed on each sample in the in-situ test sample library. Training, test, and validation data are selected from the in-situ test sample library, and an appropriate neural network data model type is chosen as the surrogate model for the first pile foundation bearing capacity. The model is then trained using the training data. Subsequently, the model performance is optimized by adjusting hyperparameters, and the model effect is evaluated using validation data, with further parameter adjustments as necessary. Finally, the final performance of the model is evaluated on an independent test set.

[0111] Figure 2 In the embodiment of the intelligent prediction method for pile foundation bearing capacity shown, before simulating each karst geological formation using the optimal geological parameters of various karst geological formations to calculate the pile foundation bearing capacity parameters of each karst geological formation and form a pile foundation simulation sample library, the method further includes:

[0112] The parameters of the third pile foundation bearing capacity surrogate model are adjusted by using the pile foundation field test sample library.

[0113] The optimal geological parameters are obtained by inversion using the third pile foundation bearing capacity proxy model.

[0114] This embodiment obtained accurate actual geological parameters through full-scale on-site construction. These parameters were used to adjust the parameters of the surrogate model for the bearing capacity of the third pile foundation, ensuring that its output matched the actual test data as closely as possible. This involved optimization algorithms and data fitting techniques. Optimal geological parameters under different working conditions were thus obtained. Simulations could be performed based on these optimal geological parameters, making the boundary conditions, cave size, and location of the finite element simulation more accurate. This ensured that the performance of the surrogate model obtained under different working conditions was consistent with the actual test data.

[0115] This invention combines experimental data, numerical simulation, and machine learning. Data from in-situ tests, model tests, and finite element simulations, including stress, strain, load, characteristic points of settlement curves, and curve shape parameters, are used as input data. This data is divided into training and test sets. The parameters are used as an input sample library to train a neural network and find the optimal parameter set. The algorithm is adaptive, adjusting model parameters according to different karst geological conditions to improve prediction accuracy and applicability. By combining extensive geological data and real-time monitoring data, machine learning algorithms are used to analyze and predict the impact of geological changes on pile foundation stability. Machine learning technology improves the reliability and prediction accuracy of pile foundation construction in karst geological environments. It reduces additional costs and delays caused by geological risks, optimizes project management and resource utilization, and provides an intelligent decision support system to help engineering teams make timely and accurate decisions in complex geological environments.

[0116] Figure 7 This is a structural schematic diagram of an intelligent prediction device for pile foundation bearing capacity provided in Embodiment 3 of the present invention. Figure 7 As shown, the device includes:

[0117] The first sample acquisition unit 701 is used to acquire geological parameters of various karst geology based on engineering in-situ tests, and to form an in-situ test sample library by taking one of the geological parameters of the karst geology as a sample.

[0118] Inversion unit 702 is used to invert the karst geology based on the trained first pile foundation bearing capacity surrogate model using the actual observation data of the karst geology, and obtain the optimal geological parameters of the karst geology.

[0119] The second sample acquisition unit 703 is used to acquire geological parameters of field tests under multiple working conditions to form a sample library of pile foundation field tests.

[0120] The third sample acquisition unit 704 is used to acquire pile bearing capacity parameters of various karst geological conditions under the pile foundation scale model experiment, and form a pile foundation scale model sample library.

[0121] The fourth sample acquisition unit 705 is used to simulate each of the karst geologies using the optimal geological parameters of various karst geologies, calculate the pile foundation bearing capacity parameters of each of the karst geologies, and form a pile foundation simulation sample library.

[0122] The fourth sample acquisition unit 706 is used to construct a pile foundation bearing capacity training sample library by combining the pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library.

[0123] Training unit 707 is used to train the pile foundation bearing capacity prediction model using the pile foundation bearing capacity training sample library;

[0124] The prediction unit 708 is used to input the geological parameters of the pile foundation model at the actual construction site into the trained pile foundation bearing capacity prediction model to obtain the optimal geological parameters;

[0125] The model optimization unit 709 is used to optimize and adjust the pile foundation model at the actual construction site based on the optimal geological parameters.

[0126] The intelligent prediction device for pile foundation bearing capacity provided in the embodiments of the present invention can execute the intelligent prediction device for pile foundation bearing capacity provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0127] Figure 8A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0128] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0129] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0130] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as an intelligent prediction method for pile foundation bearing capacity.

[0131] In some embodiments, a method for intelligent prediction of pile foundation bearing capacity may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the intelligent prediction method for pile foundation bearing capacity described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform an intelligent prediction method for pile foundation bearing capacity by any other suitable means (e.g., by means of firmware).

[0132] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0133] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0134] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0135] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0136] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0137] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0138] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0139] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for intelligent prediction of pile foundation bearing capacity, characterized in that, include: Based on in-situ engineering tests, geological parameters of various karst geological types are obtained, and an in-situ test sample library is formed by using one of the geological parameters of the karst geological type as a sample. Using the actual observation data of the karst geology, the optimal geological parameters of the karst geology are obtained by inversion based on the trained first pile foundation bearing capacity surrogate model; Geological parameters from field tests under multiple working conditions were obtained to form a sample library of field tests for pile foundations; The bearing capacity parameters of various karst geological conditions under the pile foundation scale model test were obtained to form a pile foundation scale model sample library; Using the optimal geological parameters of various karst geological formations, simulations are performed on each of the karst geological formations to calculate the pile foundation bearing capacity parameters of each karst geological formation, thus forming a pile foundation simulation sample library. The pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library are combined to construct a pile foundation bearing capacity training sample library; The pile foundation bearing capacity prediction model was trained using the aforementioned pile foundation bearing capacity training sample library; The optimal geological parameters are obtained by inputting the actual geological parameters from the construction site into the trained pile foundation bearing capacity prediction model. The pile foundation model at the actual construction site was optimized and adjusted based on the optimal geological parameters.

2. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, Also includes: The geological parameters obtained from the in-situ civil engineering test are compared with the test results of the scaled-down pile foundation model experiment; the method of the scaled-down pile foundation model experiment or the design of the pile foundation model is adjusted according to the comparison results.

3. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, Also includes: The geological parameters of various karst geological formations under reduced dimensions are substituted into the second pile foundation bearing capacity proxy model for inversion to obtain the optimal geological parameters of various karst geological formations under reduced dimensions.

4. The intelligent prediction method for pile foundation bearing capacity according to claim 3, characterized in that, Before inverting the geological parameters of various karst geological formations under reduced dimensions into the second pile foundation bearing capacity surrogate model to obtain the optimal geological parameters of various karst geological formations under reduced dimensions, the following steps are also included: Select training, test, and validation sets from the pile foundation scaled model sample library; The second pile foundation bearing capacity surrogate model was trained using the training set; The performance of the second pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the second pile foundation bearing capacity surrogate model is evaluated using the validation set. The final performance of the second pile foundation bearing capacity proxy model is evaluated using the test set.

5. The intelligent prediction method for pile foundation bearing capacity according to claim 3, characterized in that, Also includes: The optimal geological parameters of various karst geological formations under scaled-down conditions are used to simulate each karst geological formation, calculate the pile foundation bearing capacity of each karst geological formation, and obtain a scaled-down simulation sample library. The data in the scaled-down simulation sample library is then used to supplement the scaled-down pile foundation model sample library.

6. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, Also includes: By converting data from field tests of pile foundation bearing capacity in various karst geological conditions with simulation data of various karst geological conditions under the same working conditions, the simulation sample library of pile foundations obtained from the simulation is corrected.

7. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, The optimal geological parameters of the karst geology are obtained by inversion based on the trained first pile foundation bearing capacity surrogate model using actual observation data of the karst geology. This process also includes: The training set, test set, and validation set are selected from the in-situ test sample library; The first pile foundation bearing capacity proxy model was trained using the training set; The performance of the first pile foundation bearing capacity surrogate model is optimized by adjusting hyperparameters, and the effectiveness of the first pile foundation bearing capacity surrogate model is evaluated using the validation set. The final performance of the first pile foundation bearing capacity proxy model is evaluated using the test set.

8. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, Before simulating each karst geological formation using the optimal geological parameters of various karst geological formations, calculating the pile foundation bearing capacity parameters of each karst geological formation, and forming a pile foundation simulation sample library, the following steps are also included: The parameters of the third pile foundation bearing capacity surrogate model were adjusted by using the aforementioned pile foundation field test sample library. The optimal geological parameters are obtained by inversion using the third pile foundation bearing capacity proxy model.

9. The intelligent prediction method for pile foundation bearing capacity according to claim 1, characterized in that, Before training the pile foundation bearing capacity prediction model using the aforementioned pile foundation bearing capacity training sample library, the following steps are also included: Data preprocessing, noise reduction, and feature extraction are performed on the data in the pile foundation bearing capacity training sample library.

10. An intelligent prediction device for pile foundation bearing capacity, characterized in that, include: The first sample acquisition unit is used to acquire geological parameters of various karst geology based on in-situ engineering tests, and to form an in-situ test sample library by using one of the geological parameters of the karst geology as a sample. The inversion unit is used to invert the karst geology based on the trained first pile foundation bearing capacity surrogate model using the actual observation data of the karst geology, so as to obtain the optimal geological parameters of the karst geology. The second sample acquisition unit is used to acquire geological parameters of field tests under multiple working conditions to form a sample library of pile foundation field tests. The third sample acquisition unit is used to acquire pile bearing capacity parameters of various karst geological conditions under the pile foundation scale model experiment, and form a pile foundation scale model sample library. The fourth sample acquisition unit is used to simulate each of the karst geologies using the optimal geological parameters of various karst geologies, calculate the pile foundation bearing capacity parameters of each of the karst geologies, and form a pile foundation simulation sample library. The fourth sample acquisition unit is used to construct a pile foundation bearing capacity training sample library by combining the pile foundation field test sample library, the pile foundation scale model sample library, and the pile foundation simulation sample library. The training unit is used to train the pile foundation bearing capacity prediction model using the pile foundation bearing capacity training sample library. The prediction unit is used to input the geological parameters of the pile foundation model at the actual construction site into the trained pile foundation bearing capacity prediction model to obtain the optimal geological parameters; The model optimization unit is used to optimize and adjust the pile foundation model at the actual construction site based on the optimal geological parameters.

Images