CN-121997207-A - Pile body integrity intelligent detection method and system based on temperature field simulation data drive

CN121997207ACN 121997207 ACN121997207 ACN 121997207ACN-121997207-A

Abstract

The invention discloses a pile body integrity intelligent detection method and system based on temperature field simulation data driving, wherein the method comprises the steps of constructing a mixed data set formed by mixing a temperature field simulation database and a pile body integrity field detection database, wherein each data point comprises working condition parameters of a cast-in-place pile, temperature field data and pile body integrity categories; the temperature field simulation database is built through a foundation pile hydration heat finite element analysis model, a pile body integrity intelligent judgment model is obtained by training through a mixed data set based on a machine learning regression model, temperature field data and corresponding working condition parameters of the cast-in-place pile to be detected are obtained and input into the pile body integrity intelligent judgment model, and the pile body integrity category of the cast-in-place pile to be detected is output. The invention can realize early and rapid detection of pile formation by mixing training models of physical simulation and measured data based on the distribution of the hydration heat temperature field, effectively distinguish geological interference and real defects by introducing multidimensional working condition parameters, and improve detection reliability.

Inventors

Jiao shuai
WU HONGXI
LIN CHAOQUN
HUANG QIYUN
Zhong Aidi
YAO DONGMING
YANG SIYU
FENG ZIHAO

Assignees

广东省有色工业建筑质量检测站有限公司

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (10)

1. The intelligent pile body integrity detection method based on temperature field simulation data driving is characterized by comprising the following steps of: S1, constructing a mixed data set, wherein the mixed data set is formed by mixing a temperature field simulation database and a pile body integrity field detection database, and each data point in the mixed data set comprises working condition parameters of a cast-in-place pile, temperature field data and a corresponding pile body integrity category; S2, training by utilizing the mixed data set based on a machine learning regression model to obtain an intelligent pile body integrity judging model; s3, acquiring working condition parameters corresponding to the cast-in-situ pile to be detected and temperature field data of the cast-in-situ pile distributed along the depth direction of the pile body; s4, inputting the working condition parameters and the temperature field data in the step S3 into the intelligent pile body integrity judging model, and outputting the pile body integrity type of the cast-in-place pile to be tested by the intelligent pile body integrity judging model.
2. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 1, wherein the construction of the foundation pile hydration heat finite element analysis model comprises the following steps: s11, establishing a 1:1 geometric model comprising a foundation pile entity and a rock-soil entity; S12, setting geotechnical thermal data, hydration heat parameters and defect parameters for the geometric model; s13, setting boundary conditions and initial conditions based on a Fourier heat transfer law to obtain an initial finite element analysis model; S14, comparing a temperature curve output by the initial finite element analysis model with temperature data acquired by pile body integrity field detection, and carrying out iterative adjustment on the geotechnical thermal data, the hydration heat parameter, the defect parameter and the boundary condition to calibrate the initial finite element analysis model until the fitting requirement is met, so as to obtain a foundation pile hydration heat finite element analysis model: , Wherein, the Representing a mapping function, T representing temperature field data, S representing rock-soil parameters, P representing foundation pile parameters, E representing environmental parameters, Representing boundary condition parameters.
3. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 1, wherein the construction of the temperature field simulation database specifically comprises the following steps: Step 3.1, using the foundation pile hydration heat finite element analysis model, taking working condition parameters including preset defect parameters as variables, and carrying out simulation calculation under different working condition parameters in batches through an automatic script to obtain simulation data consisting of the working condition parameters [ S, P, E ] and corresponding temperature field data T; And 3.2, according to the preset defect parameters, assigning a corresponding pile body integrity class C as a label to each group of simulation data to form a temperature field simulation database consisting of data points [ S, P, E, T and C ].
4. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 1, wherein the construction of the pile body integrity field detection database specifically comprises the following steps: step 4.1, introducing a preset defect type into the test pile in an artificial way, and constructing a field test pile with a corresponding pile body integrity type; Step 4.2, acquiring readings of a temperature sensor along the longitudinal direction of the pile body after concrete pouring, acquiring temperature field data T, and synchronously recording corresponding working condition parameters [ S, P, E ] to form actual measurement data; step 4.3, verifying whether the actual pile body integrity category of the field test pile accords with the preset defect type by adopting an existing pile body integrity detection method; And 4.4, assigning a pile body integrity category C which is verified to be consistent to each group of measured data as a label to form a pile body integrity field detection database consisting of data points [ S, P, E, T and C ].
5. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 1, wherein the training of the intelligent pile body integrity judgment model comprises the following steps: S21, randomly mixing the temperature field simulation database with a pile body integrity field detection database to obtain a mixed data set, and using working condition parameters [ S, P, E ] and temperature field data T as characteristic variables and pile body integrity class C as target variables for model training; S22, determining the type of the candidate machine learning regression model and the corresponding super-parameter search range; S23, constructing a joint search space omega, wherein each candidate configuration theta of the joint search space omega is formed by a candidate machine learning regression model type and a corresponding hyper-parameter combination; s24, performing K-fold cross validation on each candidate configuration theta by using the mixed data set, and calculating an average decision coefficient and an average mean square error of each candidate configuration theta: , , Wherein K is the cross-validation fold number, For the decision coefficients of the model on the verification set at the time of verification of the kth fold, The mean square error of the model on the verification set in the k-th fold verification is obtained; s25, selecting a candidate configuration with the largest average decision coefficient as a global optimal configuration omega by taking the constraint condition that the average decision coefficient is larger than 0.95 and the average mean square error is smaller than 0.05; s26, retraining by utilizing all the mixed data sets based on the global optimal configuration omega to obtain the intelligent pile body integrity judging model.
6. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 1, wherein in step S2, the machine learning regression model comprises a linear regression model, a ridge regression model, a lasso regression model, a decision tree regression model, a random forest regression model, a support vector regression model and a neural network regression model.
7. The intelligent pile body integrity detection method based on the temperature field simulation data driving of claim 1, wherein the working condition parameters comprise rock-soil parameters, foundation pile parameters and environment parameters, wherein the rock-soil parameters comprise types and thicknesses of rock-soil layers and marks of whether the rock-soil layers are below a groundwater level or not, the foundation pile parameters comprise pile diameter, pile length, sleeve length, cement types, water-cement ratio, defect positions, defect lengths and defect types, and the environment parameters comprise initial environment temperature; The temperature field data comprise the mapping relation between the temperature of each measuring point in the hydration heat peak period and the corresponding depth; The pile body integrity type values are 1,2, 3 and 4, and correspond to the class I pile, the class II pile, the class III pile and the class IV pile respectively.
8. The intelligent pile body integrity detection method based on temperature field simulation data driving of claim 2, wherein the hydration heat parameter adopts an exponential simplification model: , Wherein Q (t) represents the total amount of hydration heat, Q 0 and m are model constant parameters related to cement types and water-cement ratios, and t is time.
9. The intelligent pile body integrity detection method based on temperature field simulation data driving according to claim 2, wherein the defect parameters comprise defect positions, defect lengths, defect types and heat conductivity coefficients or hydration heat parameters of concrete at defects according to defect type differentiation assignment.
10. The intelligent pile body integrity detection system driven by temperature field simulation data, which is applied to the method of any one of claims 1-9, is characterized by comprising the following steps: The system comprises a mixed data set construction module, a mixed data set construction module and a pile body analysis module, wherein the mixed data set is formed by mixing a temperature field simulation database and a pile body integrity field detection database, and each data point in the mixed data set comprises working condition parameters of a cast-in-place pile, temperature field data and a corresponding pile body integrity category; The model training module is used for training to obtain an intelligent pile body integrity judging model by utilizing the mixed data set based on a machine learning regression model; The data acquisition module is used for acquiring working condition parameters corresponding to the cast-in-situ pile to be detected and temperature field data of the cast-in-situ pile distributed along the depth direction of the pile body; The intelligent judging module is used for inputting the working condition parameters and the temperature field data in the data acquisition module into the pile body integrity intelligent judging model, and the pile body integrity intelligent judging model outputs the pile body integrity type of the cast-in-place pile to be tested.

Description

Pile body integrity intelligent detection method and system based on temperature field simulation data drive Technical Field The invention belongs to the technical field of foundation pile detection, and particularly relates to an intelligent pile body integrity detection method and system based on temperature field simulation data driving. Background The cast-in-place pile is widely applied to projects such as house construction, municipal administration, bridges and the like due to the advantages of high bearing capacity, flexible construction and the like. Because the construction flow of the cast-in-place pile is complex, the cast-in-place pile is influenced by various factors such as geological conditions, construction process and the like, and the defects of necking, mud clamping, segregation, honeycomb, loosening and the like are easy to occur, so that the bearing capacity of the pile body is influenced, and the service life of the pile body is prolonged. Therefore, quality control and integrity detection of the cast-in-place pile are of paramount importance. At present, the main methods for detecting the integrity of the cast-in-place pile comprise a high-strain method, a low-strain method, an acoustic wave transmission method, a core drilling method and the like. However, the above method has a disadvantage in that one of them has strict requirements on the age of pile formation and the strength of concrete. The low strain method and the sound wave transmission method need concrete strength not lower than 70% of design strength, the core drilling method and the high strain method need concrete age up to 28 days, detection is difficult to be carried out in early pile forming, and construction period is influenced. Secondly, the data analysis is still mainly judged by combining the standard with engineering experience manually, and has a certain degree of subjectivity. The detection of pile body integrity by using concrete hydration heat is a new technology developed in recent years. Hydration heat is generated during the setting of concrete, typically peaking within 48 hours. In the homogeneous equal-diameter foundation pile without defects, the temperature of measuring points at the same horizontal position with different depths is basically the same, and the temperature of the measuring points is obviously lower than that of the defect-free position at the defect positions such as fracture, crack, necking, mud clamping, cavity and the like of the pile body. Thus, foundation pile defects can be determined early in pile formation by collecting temperature field data. However, the existing hydration heat detection technology still has the defects that firstly, only a detection method is provided, the integrity type of a pile body cannot be directly judged, the standard requirement of engineering acceptance is not met, secondly, foundation pile defects and complex geological conditions are similar in temperature curve, and real defects and geological interference are difficult to distinguish. Thirdly, because the pile end defect and the normal pile end show descending trend on the temperature curve, the pile end defect is difficult to accurately judge. Disclosure of Invention The invention aims to overcome the defects and shortcomings in the prior art, and provides an intelligent pile body integrity detection method based on temperature field simulation data driving, which can realize early detection of pile forming, effectively distinguish geological interference from real defects and avoid subjectivity of artificial experience judgment. The second aim of the invention is to provide an intelligent pile body integrity detection system driven by temperature field simulation data. The invention discloses an intelligent pile body integrity detection method driven by temperature field simulation data, which comprises the following steps: S1, constructing a mixed data set, wherein the mixed data set is formed by mixing a temperature field simulation database and a pile body integrity field detection database, and each data point in the mixed data set comprises working condition parameters of a cast-in-place pile, temperature field data and a corresponding pile body integrity category; S2, training by utilizing the mixed data set based on a machine learning regression model to obtain an intelligent pile body integrity judging model; s3, acquiring working condition parameters corresponding to the cast-in-situ pile to be detected and temperature field data of the cast-in-situ pile distributed along the depth direction of the pile body; s4, inputting the working condition parameters and the temperature field data in the step S3 into the intelligent pile body integrity judging model, and outputting the pile body integrity type of the cast-in-place pile to be tested by the intelligent pile body integrity judging model. Preferably, the construction of the foundation pile hydration heat finite element analysis mod