CN-122021264-A - Intelligent analysis method and system for pile foundation buckling behavior based on transparent soil model test

CN122021264ACN 122021264 ACN122021264 ACN 122021264ACN-122021264-A

Abstract

The invention discloses a pile foundation buckling behavior intelligent analysis method and system based on a transparent soil model test, and particularly relates to the technical field of engineering parameter intelligent analysis; the method is used for solving the problem that the transparent soil test observation information cannot be effectively converted into accurate design parameters due to the fact that the simplified representation of the traditional mechanical model is not matched with the true complex constraint form of the soil body around the pile; the method comprises the steps of extracting displacement field data of soil around piles from transparent soil model test sequence images, identifying potential instability paths, directly mapping and generating quantized data representing space constraint stiffness distribution of the soil according to space-time evolution characteristics of the paths by utilizing a trained machine learning model, identifying weak and reinforced constraint areas based on the data, analyzing constraint contribution degree and system redundancy improvement potential of each quantized area through virtual parameter disturbance, and finally generating pile foundation reinforcement scheme parameters aiming at the weak areas based on the quantized indexes, so that intellectualization and refinement of reinforcement design observed from tests are realized.

Inventors

XU GUIZHONG
BIAN XIA
TAO WENBIN
QIU CHENGCHUN
LIU CHAO
CHEN JI
SUN HAO
MA FANG

Assignees

盐城工学院
河海大学
安徽交控工程集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260108

Claims (10)

1. The intelligent analysis method for the buckling behavior of the pile foundation based on the transparent soil model test is characterized by comprising the following steps: S1, acquiring sequence image data of a pile foundation buckling process in a transparent soil model test, and extracting displacement field data of a pile periphery soil body from the sequence image data; S2, based on displacement field data, identifying a space region in which displacement occurs first, continuously develops or accelerates to develop in the loading process by analyzing the space-time evolution characteristics of the displacement along with the loading process so as to determine a potential instability path; S3, mapping and generating quantized data representing space constraint stiffness distribution of the soil body around the pile through a trained machine learning model according to the space-time evolution characteristics of the potential instability path; S4, identifying a weak constraint area and a reinforced constraint area which play a leading role in pile body buckling in the soil body around the pile according to the quantized data; S5, calculating constraint contribution degrees of all the reinforced constraint areas by virtually changing constraint parameters of the reinforced constraint areas and evaluating system stability change, and calculating system redundancy promotion potential of all the weak constraint areas by virtually promoting constraint parameters of the weak constraint areas and evaluating system stability change; and S6, generating pile foundation reinforcement scheme parameters aiming at the weak constraint area based on constraint contribution degree and system redundancy promotion potential.
2. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein the steps of obtaining sequence image data of a pile foundation buckling process in the transparent soil model test, extracting displacement field data of a pile surrounding soil body from the sequence image data, and comprises the following steps: Carrying out space coordinate correction on the sequence image data through a preset calibration point; Selecting an analysis area containing a pile periphery soil body from the sequence image data, and setting a calculation grid in the analysis area; calculating displacement vectors of grid points in the grid frame by frame based on a digital image correlation method, wherein the displacement vectors comprise sizes and directions; and generating displacement field data of the soil body around the pile according to the displacement vectors of all grid points.
3. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein the determination of the potential instability path by analyzing the space-time evolution characteristics of displacement along with the loading process to identify the space region in which displacement occurs first, continues to develop or accelerates to develop in the loading process based on displacement field data comprises: Based on displacement field data, respectively calculating the moment when the displacement of each spatial position exceeds a preset threshold value for the first time in the whole loading process to form first time distribution; calculating the average change rate of the displacement of each spatial position in the whole loading process to form change rate distribution; extracting the spatial position with the earliest corresponding moment in the first time distribution, and carrying out spatial clustering by combining the spatial position with the highest corresponding change rate in the change rate distribution; And taking the communication region formed after clustering as a spatial region in which displacement occurs first, continuously develops or accelerates to develop, so as to further determine a potential instability path.
4. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein the mapping and generating of the quantized data representing the spatial constraint stiffness distribution of the soil body around the pile through the trained machine learning model according to the space-time evolution characteristics of the potential instability path comprises the following steps: Extracting displacement development time sequence and spatial position relation characteristics of each point on the path from the potential instability path to form a space-time characteristic set; Inputting the space-time feature set into a trained machine learning model; processing the space-time feature set through a trained machine learning model, and outputting soil constraint stiffness values corresponding to each point on a potential instability path; and according to the space positions of each point on the potential instability path and the corresponding soil body constraint stiffness values, combining the reference constraint stiffness values given by other areas of the soil body around the pile, which are not on the potential instability path, and generating quantized data representing the space constraint stiffness distribution of the soil body around the pile.
5. The intelligent analysis method for pile foundation buckling behavior based on transparent soil model test according to claim 4, wherein the space-time feature set is processed through a trained machine learning model, and the soil body constraint stiffness values corresponding to each point on the potential instability path are output by performing nonlinear transformation and feature fusion on the input space-time feature set through the trained machine learning model and directly mapping the space-time feature set into continuous stiffness values.
6. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein the identification of the weak constraint area and the reinforced constraint area which play a leading role in pile body buckling in the soil body around the pile according to the quantized data comprises the following steps: calculating a space constraint stiffness statistical characteristic value in a soil body area around the pile based on quantized data representing space constraint stiffness distribution of the soil body around the pile; Comparing the constraint stiffness value of each spatial position in the quantized data with the spatial constraint stiffness statistical characteristic value; classifying the spatial positions with constraint stiffness values lower than the spatial constraint stiffness statistical characteristic values, and aggregating the classification results adjacent to the spatial positions to form a weak constraint area; classifying the spatial positions with the constraint stiffness values higher than the spatial constraint stiffness statistical characteristic values, and aggregating the classification results adjacent to the spatial positions to form a reinforced constraint area.
7. The intelligent analysis method for pile foundation buckling behavior based on transparent soil model test according to claim 6, wherein the calculation of the statistical characteristic value of space constraint stiffness in the soil body area around the pile is realized by traversing constraint stiffness values of all spatial positions covered by the quantized data and calculating arithmetic mean or median values of the constraint stiffness values based on quantized data representing spatial constraint stiffness distribution of soil body around the pile.
8. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein calculating the constraint contribution of each reinforced constraint area by virtually changing the constraint parameters of the reinforced constraint area and evaluating the system stability change, and calculating the system redundancy promotion potential of each weak constraint area by virtually promoting the constraint parameters of the weak constraint area and evaluating the system stability change, comprises: Aiming at each reinforced constraint area, under the condition of maintaining constraint parameters of other areas unchanged, virtually reducing the constraint parameters of the reinforced constraint area, recalculating the buckling stability index of the pile foundation based on the adjusted integral constraint parameter distribution, and calculating the constraint contribution degree of the reinforced constraint area according to the change amplitude of the index; And aiming at each weak constraint area, under the condition of maintaining constraint parameters of other areas unchanged, virtually improving the constraint parameters of the weak constraint area, recalculating the buckling stability index of the pile foundation based on the adjusted integral constraint parameter distribution, and calculating the system redundancy improvement potential of the weak constraint area according to the index change amplitude.
9. The intelligent analysis method for pile foundation buckling behavior based on a transparent soil model test according to claim 1, wherein the generating of pile foundation reinforcement scheme parameters for weak constraint areas based on constraint contribution and system redundancy promotion potential comprises: sequencing according to the system redundancy promotion potential of each weak constraint area, and determining the priority order of reinforcement implementation; Aiming at each weak constraint area to be reinforced, determining the target constraint parameter lifting amplitude to be achieved by the weak constraint area according to the system redundancy lifting potential of the weak constraint area; And combining the constraint contribution degree of the reinforced constraint area, and adjusting the arrangement position and the scale of the reinforcing measures, so that the target constraint parameter lifting amplitude of the reinforced weak constraint area is realized, and meanwhile, the buckling stability of the integral pile foundation system is maintained or optimized, thereby generating pile foundation reinforcing scheme parameters.
10. The intelligent analysis system for pile foundation buckling behavior based on the transparent soil model test is used for realizing the intelligent analysis method for pile foundation buckling behavior based on the transparent soil model test as claimed in any one of claims 1 to 9, and is characterized by comprising the following modules: the data extraction module is used for acquiring sequence image data of a pile foundation buckling process in a transparent soil model test and extracting displacement field data of a pile periphery soil body from the sequence image data; The path determining module is used for determining a potential instability path by analyzing the space-time evolution characteristics of displacement along with the loading process and identifying a space region in which displacement occurs first, continues to develop or accelerates to develop in the loading process based on the displacement field data; The data generation module is used for mapping and generating quantized data representing space constraint stiffness distribution of the soil body around the pile through a trained machine learning model according to the space-time evolution characteristics of the potential instability path; The region identification module is used for identifying weak constraint regions and reinforced constraint regions which play a leading role in pile body buckling in the soil body around the pile according to the quantized data; The virtual analysis module is used for virtually changing constraint parameters of the reinforced constraint areas and evaluating system stability changes to calculate constraint contribution of each reinforced constraint area, and virtually improving constraint parameters of the weak constraint areas and evaluating system stability changes to calculate system redundancy improvement potential of each weak constraint area; and the parameter generation module is used for generating pile foundation reinforcement scheme parameters aiming at the weak constraint area based on constraint contribution and system redundancy boosting potential.

Description

Intelligent analysis method and system for pile foundation buckling behavior based on transparent soil model test Technical Field The invention relates to the technical field of intelligent analysis of engineering parameters, in particular to an intelligent analysis method and an intelligent analysis system of pile foundation buckling behavior based on a transparent soil model test. Background In pile foundation engineering, especially in pile foundation design and reinforcement under deep pile or complex soil layer condition, accurate prediction and evaluation of buckling stability of pile body is important, and physical model test is an important means for researching pile-soil interaction and buckling mechanism. The transparent soil model test technology is applied to research of pile foundation buckling behaviors because the technology can realize non-invasive and visual observation of the deformation field in the soil body around the pile. The technology can obtain the sequential image data of pile body displacement and surrounding soil body displacement fields in the buckling process through laser slicing and image acquisition. Based on these image data, existing analysis methods rely primarily on qualitative descriptions of displacement field morphology or on fitting inversion of a macroscopically loaded-displaced curve measured by a test with a classical mechanical model, such as a Winkler foundation model, to evaluate the overall stability of the pile foundation and guide the reinforcement design. However, the existing analysis method based on the transparent soil test has the limitation that the classical mechanical model simplifies the lateral constraint action of the soil body on the pile body into a linear spring system which is continuous, uniform or distributed according to a simple rule. The highly simplified representation mode has an essential representation gap between discontinuous, non-uniform and highly nonlinear complex constraint forms and failure modes which are actually observed by a transparent soil test, particularly by a pile periphery soil body after local reinforcement treatment, so that high-dimensional and spatial visual constraint information obtained from the transparent soil test cannot be effectively converted into parameters which can be used for accurate mechanical analysis and quantitative design, and the design and effect evaluation of a reinforcement scheme still depend on rough estimation of experience judgment and macroscopic parameters to a great extent, so that the fine design and optimization based on a real constraint mechanism are difficult to realize. Disclosure of Invention In order to overcome the defects in the prior art, the invention provides an intelligent analysis method and an intelligent analysis system for pile foundation buckling behavior based on a transparent soil model test, which are used for solving the problems in the background art. In order to achieve the above purpose, the present invention provides the following technical solutions: The intelligent analysis method for the buckling behavior of the pile foundation based on the transparent soil model test comprises the following steps: S1, acquiring sequence image data of a pile foundation buckling process in a transparent soil model test, and extracting displacement field data of a pile periphery soil body from the sequence image data; S2, based on displacement field data, identifying a space region in which displacement occurs first, continuously develops or accelerates to develop in the loading process by analyzing the space-time evolution characteristics of the displacement along with the loading process so as to determine a potential instability path; S3, mapping and generating quantized data representing space constraint stiffness distribution of the soil body around the pile through a trained machine learning model according to the space-time evolution characteristics of the potential instability path; S4, identifying a weak constraint area and a reinforced constraint area which play a leading role in pile body buckling in the soil body around the pile according to the quantized data; S5, calculating constraint contribution degrees of all the reinforced constraint areas by virtually changing constraint parameters of the reinforced constraint areas and evaluating system stability change, and calculating system redundancy promotion potential of all the weak constraint areas by virtually promoting constraint parameters of the weak constraint areas and evaluating system stability change; and S6, generating pile foundation reinforcement scheme parameters aiming at the weak constraint area based on constraint contribution degree and system redundancy promotion potential. Further, obtaining sequence image data of a pile foundation buckling process in a transparent soil model test, and extracting displacement field data of a pile periphery soil body from the sequence image data, wherein the