System and method for predicting ultimate bearing capacity of post-grouting pile based on machine learning

Abstract

The invention discloses a system and a method for predicting ultimate bearing capacity of a post-grouting pile based on machine learning, and relates to the field of cast-in-situ bored pile monitoring, the prediction method comprises the following steps: collecting information of the post-grouting pile, and establishing a data set of a machine learning model; processing the data set, and performing machine learning training according to different grouting type data to obtain a corresponding prediction model; real-time measurement data of the post-grouting pile are collected through target equipment, and input information is formed; inputting the input information into a prediction system, automatically matching a machine learning prediction model, and predicting the ultimate bearing capacity of the post-grouting pile. The invention provides a scientific, accurate, convenient and efficient tool for predicting the ultimate bearing capacity of the post-grouting pile, the engineering cost is reduced, and the engineering efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of bored cast-in-place pile monitoring, and in particular to a system and method for predicting the ultimate bearing capacity of a post-grouting pile based on machine learning. Background Art

[0002] In the field of civil engineering, bored cast-in-place piles are an important form of foundation structure. The prediction of the ultimate bearing capacity of post-grouting piles is crucial to ensuring the safety and quality of the project.

[0003] Traditional prediction methods for the ultimate bearing capacity of post-grouting piles, such as theoretical calculation method, empirical formula method, traditional in-situ testing method, etc., all have obvious limitations when faced with complex geological conditions and construction processes.

[0004] Theoretical calculation method, based on soil mechanics and elastic mechanics theory, calculates the bearing capacity of a single pile through a complex mathematical model. However, due to the variability of geological conditions and the complexity of pile-soil interaction, theoretical calculations often fail to fully consider various influencing factors, resulting in large deviations between the calculation results and the actual results. In particular, in the application of post-grouting pile technology, the bearing capacity of the pile is significantly improved by strengthening the combination of the pile body and the surrounding soil through grouting, but this improvement effect is difficult to accurately quantify in theoretical calculations.

[0005] The empirical formula method is an empirical formula summarized based on a large number of engineering practices. Although it has certain practicality, it lacks adaptability to different geological conditions and construction processes. When facing the post-grouting pile technology, the empirical formula is also difficult to accurately reflect its bearing capacity improvement effect, resulting in uncertainty and error in the prediction results.

[0006] Traditional in-situ testing methods, such as static load testing, have reliable results, but data processing is time-consuming, costly, and time-consuming, and causes certain damage to the pile body, limiting their widespread application.

[0007] Therefore, in the field of bored cast-in-place pile monitoring technology, the existing technology for predicting the ultimate bearing capacity of post-grouting piles has the following defects: inaccurate prediction, complex prediction process, high cost, time-consuming data processing, etc. Summary of the invention

[0008] In view of the defects existing in the prior art, the purpose of the present invention is to provide a system and method for predicting the ultimate bearing capacity of post-grouting piles based on machine learning, providing a more scientific, accurate, convenient and efficient tool for predicting the ultimate bearing capacity of post-grouting piles, reducing engineering costs and improving engineering efficiency.

[0009] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0010] In a first aspect, an embodiment of the present application provides a prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning, the prediction system comprising: a model training module, a data acquisition module, a model prediction module, and a result display module, wherein:

[0011] The model training module is used to: construct an ultimate bearing capacity prediction model based on data and machine learning models to predict the ultimate bearing capacity of post-grouting piles under different grouting types;

[0012] The data acquisition module is used to: measure and collect information of the post-grouting piles in real time through the target device to obtain real-time measurement data;

[0013] The model prediction module is used to: predict the ultimate bearing capacity of the post-grouting pile in real time based on the real-time measurement data through a preset machine learning algorithm to obtain a prediction result;

[0014] The result display module is used to: display the prediction results for user selection, and add the prediction results to the prediction model to optimize the prediction model.

[0015] In combination with the first aspect, in one implementation, the model training module includes: a data collection submodule, a feature selection submodule, and a model training optimization submodule, wherein:

[0016] The data collection submodule is used to: collect information of post-grouting piles;

[0017] The feature selection submodule is used to: analyze the correlation between various parameters and the ultimate bearing capacity of post-grouting piles through statistical analysis and machine learning technology, identify the features that affect the prediction results, and perform engineering treatment on the selected features;

[0018] The model training and optimization submodule is used to train and optimize the prediction model through the random forest algorithm and the Bayesian optimization algorithm.

[0019] In combination with the first aspect, in one implementation, the model training module further includes: a model testing and verification submodule, a collection and prediction submodule, and a feedback optimization submodule, wherein:

[0020] The model test verification submodule is used to: verify the prediction model using test data independent of the training set, evaluate the prediction accuracy and stability of the prediction model, and verify the reliability and generalization ability of the prediction model by comparing the prediction results with the actual engineering data;

[0021] The acquisition prediction submodule is used to: obtain parameters in the post-grouting pile construction process in real time, and predict the ultimate bearing capacity of the post-grouting pile using the trained prediction model;

[0022] The feedback optimization submodule is used to: feed back the prediction results and add the prediction results to the prediction model to optimize the prediction model.

[0023] In combination with the first aspect, in one embodiment, the engineering processing of the selected features includes: encoding categorical variables into numerical values, and standardizing or normalizing continuous variables to improve the learning efficiency and prediction accuracy of the prediction model.

[0024] In combination with the first aspect, in one embodiment, the information of the post-grouting piles includes: pile diameter, pile length, bearing layer type, pile top settlement, pile end grouting cement dosage, pile end termination grouting pressure, pile side termination grouting pressure, and side end grouting cement dosage.

[0025] In a second aspect, an embodiment of the present application provides a method for predicting the ultimate bearing capacity of a post-grouting pile based on machine learning, which is applied to the prediction system described in the first aspect, comprising the following steps:

[0026] Collect information about post-grouting piles and build a data set for machine learning models;

[0027] Process the data set, conduct machine learning training based on different grouting type data, and obtain the corresponding prediction model;

[0028] The real-time measurement data of the post-grouting piles is collected by the target device to form input information;

[0029] The input information is fed into the prediction system and the machine learning algorithm is automatically matched to predict the ultimate bearing capacity of the post-grouting piles.

[0030] In combination with the second aspect, in one embodiment, the data set for establishing the machine learning model includes the following steps: organizing the data into a structured format to form a data set suitable for machine learning model training, the data set containing multiple feature variables and their corresponding ultimate bearing capacity labels to facilitate subsequent data processing and model training.

[0031] In conjunction with the second aspect, in one implementation, performing machine learning training according to different grouting type data to obtain a corresponding prediction model includes the following steps:

[0032] For the grouting type data of "end pile grouting", select "pile diameter", "pile length", "ultimate end resistance", "pile top settlement", "pile end grouting cement consumption" and "pile end grouting termination pressure" as input features;

[0033] For the grouting type data of "combined grouting", select: "Pile diameter", "Pile length", "Bearing layer type", "Pile top settlement", "Pile end pressure drop cement consumption", "Pile end grouting pressure", "Pile side grouting pressure", "Side end grouting cement consumption" as input features;

[0034] The categorical variable “bearing layer category” is encoded as a numerical type, and the method used is one-hot encoding, while the method for processing continuous variables is normal standardization.

[0035] In a third aspect, an embodiment of the present application provides a prediction device for the ultimate bearing capacity of post-grouting piles based on machine learning, the prediction device comprising a processor, a memory, and a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning stored in the memory and executable by the processor, wherein when the prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning is executed by the processor, the steps of the prediction method described in the second aspect are implemented.

[0036] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which is stored a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning, wherein when the prediction program is executed by a processor, the steps of the prediction method described in the second aspect are implemented.

[0037] Compared with the prior art, the advantages of the present invention are:

[0038] (1) Compared with the traditional prediction methods that rely on complex physical models or a large number of field tests, the present invention greatly simplifies the process. Only relevant data parameters need to be input to quickly obtain the prediction results, which not only significantly reduces time and labor costs, but also reduces engineering costs. The present invention provides engineers with a more scientific, accurate, convenient and efficient prediction tool, reduces engineering costs, improves engineering efficiency, and can achieve rapid and accurate prediction and continuous optimization of the ultimate bearing capacity of post-grouting piles under different geological conditions and construction processes, which has application value.

[0039] (2) The prediction method of the present invention has wide applicability and can adapt to different geological conditions, pile type designs, construction processes and material selections, and provide relatively accurate prediction results. This powerful generalization ability makes the present invention have great application potential and promotion value in pile foundation engineering. The present invention not only overcomes the limitations of traditional methods, but also lays a solid foundation for the further promotion and application of post-grouting pile technology, playing a more important role in civil engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0041] Figure 1 It is a structural block diagram of a system for predicting the ultimate bearing capacity of post-grouting piles based on machine learning in an embodiment of the present invention.

[0042] Figure 2 4 is a structural block diagram of a model training module in an embodiment of the present invention.

[0043] Figure 3 Flow chart of a method for predicting the ultimate bearing capacity of post-grouting piles based on machine learning in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0045] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0046] Example 1: Prediction system of ultimate bearing capacity of post-grouting piles based on machine learning

[0047] See also Figure 1 As shown, an embodiment of the present invention provides a prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning, including: a model training module, a data acquisition module, a model prediction module, and a result display module, wherein:

[0048] The model training module is used to: construct a machine learning-based prediction model for the ultimate bearing capacity of post-grouting piles based on a large amount of data and machine learning models. This prediction model can be used to predict the ultimate bearing capacity under different grouting types.

[0049] The data acquisition module is used to: measure and collect relevant information of the post-grouting piles on site in real time through the target equipment to obtain real-time measurement data;

[0050] The relevant information of post-grouting piles includes pile diameter, pile length, bearing layer type, pile top settlement, pile end grouting cement consumption, pile end grouting termination pressure, pile side grouting termination pressure, and side end grouting cement consumption.

[0051] The model prediction module is used to: based on real-time measurement data, through a preset machine learning algorithm, make real-time predictions on the ultimate bearing capacity of the post-grouting piles to obtain prediction results.

[0052] The result display module is used to: intuitively display the prediction results for users to choose from, and add the prediction results to the model training module to further enrich and optimize the training model and continuously improve the prediction accuracy.

[0053] Example 2: Model training module

[0054] See also Figure 2 As shown, the model training module in the embodiment of the present invention specifically includes: a data collection submodule, a feature selection submodule, a model training optimization submodule, a model test verification submodule, a collection prediction submodule, and a feedback optimization submodule, wherein:

[0055] The data collection submodule is used to collect data related to the ultimate bearing capacity of post-grouting piles from existing engineering databases and literature, including but not limited to pile diameter, pile length, bearing layer type, pile top settlement, pile end grouting cement consumption, pile end termination grouting pressure, pile side termination grouting pressure and side end grouting cement consumption, etc.

[0056] The feature selection submodule is used to: deeply analyze the correlation between various parameters and the ultimate bearing capacity of post-grouting piles through statistical analysis and machine learning techniques, identify the features that have a significant impact on the prediction results, and perform necessary engineering processing on the selected features, such as encoding categorical variables into numerical types and standardizing or normalizing continuous variables, so as to improve the learning efficiency and prediction accuracy of the model.

[0057] The model training optimization submodule is used to: use the Bayesian optimization algorithm to optimize the random forest model for training, and combine the known ultimate bearing capacity data to improve the model's predictive ability.

[0058] After comparing two different grouting types, the machine learning method finally selected was the random forest algorithm, and the optimization algorithm was the Bayesian optimization algorithm.

[0059] In the model training and optimization submodule,

[0060] The parameters of the random forest algorithm for the prediction of “end pile grouting” are:

[0061] bootstrap='True', max_depth=15, max_features='auto', min_samples_leaf=3, min_samples_split=2, n_estimators=14.

[0062] The random forest algorithm for “combined grouting” prediction is:

[0063] bootstrap='False', max_depth=20, max_features='log2', min_samples_leaf=1, min_samples_split=3, n_estimators=15.

[0064] "bootstrap" is the self-service sampling method, "max_depth" is the maximum depth of the tree, "max_features" is the number of automatically selected features, "min_samples_leaf" is the minimum number of samples for leaf nodes, "min_samples_split" is the minimum number of samples required to split internal nodes, and "n_estimators" is the number of trees.

[0065] The model test and verification submodule is used to: verify the model using test data independent of the training set, evaluate the prediction accuracy and stability of the model, and verify the reliability and generalization ability of the model by comparing the prediction results with the actual engineering data.

[0066] The data collection and prediction submodule is used to: At the construction site, through high-precision sensors and automated data acquisition systems, obtain the key parameters of the post-grouting pile construction process in real time, and use the trained machine learning model to quickly and accurately predict the ultimate bearing capacity of the post-grouting piles, providing a scientific basis for engineering design and construction.

[0067] The feedback optimization submodule is used to: provide timely feedback of prediction results to engineering personnel, incorporate data into the database according to actual conditions, and continuously enrich the database to optimize and improve the prediction model, thereby further improving the prediction accuracy and applicability.

[0068] Example 3: Prediction method of ultimate bearing capacity of post-grouting piles based on machine learning

[0069] See also Figure 3 As shown, an embodiment of the present invention provides a method for predicting the ultimate bearing capacity of a post-grouting pile based on machine learning, comprising the following steps:

[0070] S1. Collect data and establish data set: Collect basic data information of post-grouting piles from the literature database, and establish the data set of machine learning model based on this;

[0071] S2, model training: process the data set, distinguish different grouting type data, and perform machine learning training to obtain the corresponding optimal model;

[0072] S3. Collect key parameters and real-time measurement data of the post-grouting pile construction process through the target equipment, such as grouting pressure, grouting volume, pile material characteristics and geological conditions, to form comprehensive input information;

[0073] S4. Input the comprehensive input information into the prediction system, automatically match the appropriate machine learning algorithm, and predict the ultimate bearing capacity of the post-grouting piles, providing an efficient and accurate solution for the ultimate bearing capacity prediction of the post-grouting technology.

[0074] The execution subject of the embodiment of the present invention may be:

[0075] The prediction system of the ultimate bearing capacity of post-grouting piles based on machine learning in Example 1,

[0076] The execution subject may also be a terminal or a server, which is not specifically limited here.

[0077] The embodiment of the present invention is described by taking the prediction system of the ultimate bearing capacity of post-grouting piles based on machine learning in Example 1 as the execution body.

[0078] Example 4: Specific implementation of step S1

[0079] Step S1, collecting data and establishing a data set: collecting basic data information of post-grouting cast-in-place piles from the literature database, and establishing a data set for the machine learning model based on this.

[0080] In the embodiment of the present invention, step S1 specifically includes the following steps:

[0081] Data collection: Through extensive review and screening of literature databases in the relevant civil engineering field, detailed data information related to the ultimate bearing capacity of post-grouting cast-in-place piles is systematically collected and organized.

[0082] These data include but are not limited to:

[0083] Design parameters of the pile, such as pile diameter and pile length;

[0084] Construction conditions, such as grouting process and grouting time;

[0085] Geological conditions, such as soil layer distribution and bearing layer type;

[0086] The corresponding measured data of ultimate bearing capacity, such as: pile top settlement, pile end grouting cement consumption, pile end grouting termination pressure, pile side grouting termination pressure and side end grouting cement consumption, etc.

[0087] The data collected in the embodiments of the present invention must be strictly screened and verified to ensure its accuracy and reliability.

[0088] After the data is collected, the raw data is cleaned to remove errors, duplications or invalid data to ensure the quality and integrity of the data. This process involves steps such as data missing value processing, outlier detection and correction.

[0089] Establish a data set: Organize these data into a structured format to form a data set suitable for machine learning model training. The data set should contain multiple feature variables and their corresponding ultimate bearing capacity labels to facilitate subsequent data processing and model training.

[0090] Example 5: Specific implementation of step S2

[0091] Step S2, model training: Process the data set, distinguish different grouting type data, and perform machine learning training to obtain the corresponding optimal prediction model.

[0092] In the embodiment of the present invention, step S2 specifically includes the following steps:

[0093] Data preprocessing: Perform necessary preprocessing on the data set, including data cleaning, standardization, and normalization, to eliminate noise and ensure data consistency.

[0094] For continuous variables, such as pile diameter, pile length, grouting volume, etc., the normal standardization method is used to eliminate the impact of dimensional differences on model training.

[0095] For categorical variables, such as the bearing layer category, the One-Hot Encoding method is used to convert it into numerical features for easy processing by machine learning algorithms.

[0096] Data classification: According to the different grouting types, for example, "end pile grouting" type and "combined grouting" type, the data set is divided into two independent subsets. Furthermore, each subset is divided into a training set and a test set in a ratio of 8:2 to ensure the effectiveness and generalization ability of model training.

[0097] Feature selection: In-depth analysis of the correlation between various parameters and the ultimate bearing capacity of post-grouting piles, and identification of the features that have the most significant impact on the prediction results through statistical analysis and machine learning techniques.

[0098] Perform necessary transformations on the selected features, such as encoding categorical variables into numerical values ​​and standardizing or normalizing continuous variables, to improve the learning efficiency and prediction accuracy of the model.

[0099] In the process of feature selection, different feature variables are selected for different grouting types.

[0100] For example, for the grouting type data of "end pile grouting", you can select: "pile diameter", "pile length", "ultimate end resistance", "pile top settlement", "pile end grouting cement consumption" and "pile end termination grouting pressure" as input features.

[0101] For the grouting type data of "combined grouting", you can select: "Pile diameter", "Pile length", "Bearing layer category", "Pile top settlement", "Pile end pressure drop cement consumption", "Pile end termination grouting pressure", "Pile side termination grouting pressure", and "Side end grouting cement consumption" as input features.

[0102] The categorical variable “bearing layer category” is encoded as a numerical type, and the method used is one-hot encoding, while the method for processing continuous variables is normal standardization.

[0103] Model training and optimization: According to the characteristics of the problem, select appropriate machine learning algorithms, such as random forest, GBDT (Gradient Boosting Decision Tree), neural network, etc., to build a prediction model.

[0104] Through cross-validation, grid search, Bayesian optimization and other strategies, the model's hyperparameters are fine-tuned to obtain the best training results and prediction performance.

[0105] In the model training and optimization steps, after comparing two different grouting types, the machine learning method finally selected was the random forest algorithm, and the optimization algorithm was the Bayesian optimization algorithm.

[0106] In the steps of model training and optimization,

[0107] The parameters of the random forest algorithm for the prediction of “end pile grouting” are:

[0108] bootstrap='True', max_depth=15, max_features='auto', min_samples_leaf=3, min_samples_split=2, n_estimators=14.

[0109] The random forest algorithm for “combined grouting” prediction is:

[0110] bootstrap='False', max_depth=20, max_features='log2', min_samples_leaf=1, min_samples_split=3, n_estimators=15.

[0111] "bootstrap" is the self-service sampling method, "max_depth" is the maximum depth of the tree, "max_features" is the number of automatically selected features, "min_samples_leaf" is the minimum number of samples for leaf nodes, "min_samples_split" is the minimum number of samples required to split internal nodes, and "n_estimators" is the number of trees.

[0112] Model selection and training: After comparative analysis, the random forest algorithm was finally selected as the machine learning algorithm in the embodiment of the present invention. The random forest algorithm performs well in processing complex data sets due to its good anti-overfitting ability and high prediction accuracy.

[0113] In the model comparison analysis, the embodiment of the present invention compares: ridge regression, deep learning, random forest regression, and distributed gradient boosting tree regression, among which the random forest algorithm performs best.

[0114] Different random forest model parameters are set for different grouting type data. For example, "bootstrap" is the self-service sampling method, "max_depth" is the maximum depth of the tree, "max_features" is the number of automatically selected features, "min_samples_leaf" is the minimum number of samples for leaf nodes, "min_samples_split" is the minimum number of samples required to split internal nodes, and "n_estimators" is the number of trees.

[0115] For the "end pile grouting" type data, set bootstrap to True, max_depth to 15, max_features to "auto", min_samples_leaf to 3, min_samples_split to 2, and n_estimators to 14.

[0116] For the "combined grouting" type data, adjust bootstrap to False, max_depth to 20, max_features to log2, min_samples_leaf to 1, min_samples_split to 3, and n_estimators to 15.

[0117] The selection of these parameters is based on the results of multiple experiments and Bayesian optimization algorithms, aiming to achieve the best model performance.

[0118] Model testing and validation: Use test data independent of the training set to validate the model and evaluate the model's prediction accuracy and stability. Verify the reliability and generalization ability of the model by comparing the prediction results with actual engineering data.

[0119] Real-time data collection and prediction: At the construction site, high-precision sensors and automated data collection systems are used to obtain key parameters of the post-grouting pile construction process in real time, such as grouting pressure, grouting volume, pile material properties, and geological conditions.

[0120] Based on these real-time measurement data, the ultimate bearing capacity of post-grouting piles can be quickly and accurately predicted using the trained machine learning model, providing a scientific basis for engineering design and construction.

[0121] In the step of real-time data collection and prediction, various key parameters in the construction process of post-grouting piles are collected in real time, such as grouting pressure, ultimate end resistance, cement consumption for grouting at the pile end, pile top settlement, etc. These data are directly input into the prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning in Example 1 of the present invention. The prediction system automatically processes the data and matches it with the most suitable trained machine learning model to realize online prediction of the ultimate bearing capacity of post-grouting piles, and displays the prediction results in real time on the monitoring interface for engineering personnel to immediately refer to and adjust the construction plan.

[0122] Feedback and optimization of prediction results: Feedback the prediction results to engineering personnel in a timely manner, incorporate the data into the database according to the actual situation, and continuously enrich the database to optimize and improve the prediction model to further improve the prediction accuracy and applicability.

[0123] Feedback and optimization of prediction results include the following steps:

[0124] First, the prediction results are fed back to engineering personnel in a timely and accurate manner so that they can adjust the construction plan or take corresponding measures based on the prediction information.

[0125] At the same time, the monitoring data from the actual project are compared and analyzed with the prediction results to evaluate the accuracy and reliability of the prediction model.

[0126] These data are incorporated into the database as new training samples or used to adjust model parameters, thereby continuously enriching the content of the database and enhancing the generalization ability and prediction accuracy of the model.

[0127] In addition, the performance of the model is evaluated regularly, and the model is iteratively optimized according to engineering needs and technological development to further improve its applicability and prediction effect.

[0128] Example 6: Specific implementation of step S3

[0129] Step S3, collecting key parameters in the post-grouting pile construction process through the target device, such as grouting pressure, grouting volume, pile body material properties and geological conditions, etc., to form comprehensive input information;

[0130] In the embodiment of the present invention, step S3 specifically includes the following steps:

[0131] Selection parameters: Determine the key parameters that have a significant impact on the ultimate bearing capacity during the construction of post-grouting piles, such as grouting pressure, grouting volume, pile material properties, geological conditions, etc.

[0132] Collect data: Use specialized target equipment to collect data on these key parameters in real time at the construction site to ensure data accuracy and completeness.

[0133] Integrate data: Integrate the collected data into a comprehensive set of input information for subsequent ultimate bearing capacity prediction.

[0134] Example 7: Specific implementation of step S4

[0135] Step S4: input the real-time measurement data and comprehensive input information into the prediction system, automatically match the appropriate machine learning algorithm, and predict the ultimate bearing capacity of the post-grouting piles, providing an efficient and accurate solution for the ultimate bearing capacity prediction of the post-grouting technology.

[0136] In the embodiment of the present invention, step S4 specifically includes the following steps:

[0137] Model deployment: Deploy the trained optimal random forest model to the prediction system, which can be a server, terminal or other computing device.

[0138] The prediction system needs to have efficient data processing capabilities and a stable operating environment to ensure the accuracy and reliability of real-time predictions.

[0139] Real-time prediction: The data collected in step S3 and the comprehensive input information are input into the prediction system, which will automatically match the appropriate machine learning algorithm, i.e. the deployed random forest model, and quickly calculate the ultimate bearing capacity prediction results of the post-grouting piles based on the input information. The prediction results will be fed back to the engineers in real time, providing a scientific basis for engineering design and construction.

[0140] Application of results: Engineers can use the prediction results to more accurately evaluate the bearing capacity of post-grouting piles, thereby optimizing engineering design and construction plans.

[0141] At the same time, the prediction results can also be used as monitoring indicators during the construction process to detect potential problems in a timely manner and take corresponding measures.

[0142] In addition, as the construction process progresses and data continues to accumulate, the model can be regularly updated and optimized to improve prediction accuracy and generalization capabilities.

[0143] Example 8: Prediction device for ultimate bearing capacity of post-grouting piles based on machine learning

[0144] An embodiment of the present invention provides a prediction device for the ultimate bearing capacity of post-grouting piles based on machine learning, comprising a processor, a memory, and a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning stored in the memory and executable by the processor, wherein the prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning, when executed by the processor, implements the steps of the prediction method in any one of embodiments 3 to 7, which will not be repeated here.

[0145] The prediction device can be a device with data processing function, such as a personal computer (PC), a notebook computer, or a server.

[0146] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0147] Embodiment 9: A computer-readable storage medium

[0148] An embodiment of the present invention also provides a computer-readable storage medium, on which is stored a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning, wherein the prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning, when executed by a processor, implements the steps of the prediction method in any one of Embodiments 3 to 7, which will not be repeated here.

[0149] It should be noted that the serial numbers of the above-mentioned embodiments of the present application are only for description and do not represent the advantages or disadvantages of the embodiments.

[0150] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit "first", "second" and "third" to different types.

[0151] In the description of the embodiments of the present application, "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a specific way.

[0152] In the description of the embodiments of the present application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; the “and / or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, “multiple” refers to two or more than two.

[0153] In some processes described in the embodiments of the present application, multiple operations or steps that appear in a specific order are included, but it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or in parallel, and the sequence number of the operation is only used to distinguish the different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in sequence or in parallel, and these operations or steps may be combined.

[0154] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD) as described above, and includes a number of instructions for a terminal device to execute the methods described in each embodiment of the present application.

[0155] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning, characterized in that: The prediction system includes: model training module, data collection module, model prediction module, and result display module, among which: The model training module is used to: construct an ultimate bearing capacity prediction model based on data and machine learning models to predict the ultimate bearing capacity of post-grouting piles under different grouting types; The data acquisition module is used to: measure and collect information of the post-grouting piles in real time through the target device to obtain real-time measurement data; The model prediction module is used to: predict the ultimate bearing capacity of the post-grouting pile in real time based on the real-time measurement data through a preset machine learning algorithm to obtain a prediction result; The result display module is used to: display the prediction results for user selection, and add the prediction results to the prediction model to optimize the prediction model.

2. The prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning according to claim 1, characterized in that: The model training module includes: a data collection submodule, a feature selection submodule, and a model training optimization submodule, wherein: The data collection submodule is used to: collect information of post-grouting piles; The feature selection submodule is used to: analyze the correlation between various parameters and the ultimate bearing capacity of post-grouting piles through statistical analysis and machine learning technology, identify the features that affect the prediction results, and perform engineering treatment on the selected features; The model training and optimization submodule is used to train and optimize the prediction model through the random forest algorithm and the Bayesian optimization algorithm.

3. The prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning as claimed in claim 2, characterized in that: The model training module also includes: a model testing and verification submodule, a collection and prediction submodule, and a feedback optimization submodule, wherein: The model test verification submodule is used to: verify the prediction model using test data independent of the training set, evaluate the prediction accuracy and stability of the prediction model, and verify the reliability and generalization ability of the prediction model by comparing the prediction results with the actual engineering data; The acquisition prediction submodule is used to: obtain parameters in the post-grouting pile construction process in real time, and predict the ultimate bearing capacity of the post-grouting pile using the trained prediction model; The feedback optimization submodule is used to: feed back the prediction results and add the prediction results to the prediction model to optimize the prediction model.

4. The prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning as claimed in claim 2, characterized in that: The engineering processing of the selected features includes: encoding categorical variables into numerical values ​​and standardizing or normalizing continuous variables to improve the learning efficiency and prediction accuracy of the prediction model.

5. The prediction system for the ultimate bearing capacity of post-grouting piles based on machine learning as claimed in claim 2, characterized in that: The information of the post-grouting piles includes: pile diameter, pile length, bearing layer type, pile top settlement, pile end grouting cement consumption, pile end grouting termination pressure, pile side grouting termination pressure, and side end grouting cement consumption.

6. A method for predicting the ultimate bearing capacity of post-grouting piles based on machine learning applied to the prediction system according to any one of claims 1 to 5, characterized in that: The prediction method includes the following steps: Collect information about post-grouting piles and build a data set for machine learning models; Process the data set, conduct machine learning training based on different grouting type data, and obtain the corresponding prediction model; The real-time measurement data of the post-grouting piles is collected by the target device to form input information; The input information is fed into the prediction system and the machine learning algorithm is automatically matched to predict the ultimate bearing capacity of the post-grouting piles.

7. The method for predicting the ultimate bearing capacity of post-grouting piles based on machine learning according to claim 6, characterized in that: The method for establishing a data set for a machine learning model includes the following steps: organizing the data into a structured format to form a data set suitable for machine learning model training, wherein the data set includes multiple feature variables and their corresponding ultimate bearing capacity labels to facilitate subsequent data processing and model training.

8. The method for predicting the ultimate bearing capacity of post-grouting piles based on machine learning according to claim 6, characterized in that: The method of performing machine learning training according to different grouting type data to obtain a corresponding prediction model includes the following steps: For the grouting type data of "End pile grouting", select "Pile diameter", "Pile length", "Ultimate end resistance", "Pile top settlement", "Pile end grouting cement consumption" and "Pile end grouting termination pressure" as input features; For the grouting type data of "Combined grouting", select "Pile diameter", "Pile length", "Bearing layer type", "Pile top settlement", "Pile end pressure drop cement consumption", "Pile end grouting pressure", "Pile side grouting pressure", and "Side end grouting cement consumption" as input features; The categorical variable "bearing layer category" is encoded as a numerical type, and the method used is one-hot encoding. The method for processing continuous variables is normal standardization.

9. A prediction device for the ultimate bearing capacity of post-grouting piles based on machine learning, characterized in that: The prediction device includes a processor, a memory, and a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning stored in the memory and executable by the processor, wherein when the prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning is executed by the processor, the steps of the prediction method as described in any one of claims 6 to 8 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a prediction program for the ultimate bearing capacity of post-grouting piles based on machine learning, wherein when the prediction program is executed by a processor, the steps of the prediction method as described in any one of claims 6 to 8 are implemented.

Images