CN-121997781-A - Dynamic monitoring method, system and equipment for roadbed rebound modulus

CN121997781ACN 121997781 ACN121997781 ACN 121997781ACN-121997781-A

Abstract

The application relates to the technical field of rebound modulus monitoring, in particular to a dynamic monitoring method, a system and equipment for roadbed rebound modulus, wherein the method comprises the steps of preparing a plurality of groups of soil body samples aiming at each working condition, respectively carrying out multistage loading dynamic triaxial test on the soil body samples to generate hysteresis curves of various strain grades; the method comprises the steps of obtaining discrete coefficients of soil samples at each strain level, correcting initial dynamic elastic modulus extracted from hysteresis curves, building a nonlinear constitutive model responding to soil state change, generating a simulation data set containing a plurality of feature vectors and rebound modulus thereof, training an integrated learning model, and estimating compaction quality of a roadbed by utilizing the integrated learning model to predict the rebound modulus according to the obtained feature vectors on a roadbed construction site. The method improves the accuracy of monitoring the road base rebound modulus in real time.

Inventors

QIAN JINSONG
Ouyang Huiyi
LIU JIAN
ZHOU TONG
CHEN XINRAN
CHEN GUANG
BAI YANG

Assignees

同济大学

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. A method for dynamically monitoring the rebound modulus of a roadbed, which is characterized by comprising the following steps: Controlling physical parameters and mechanical environment parameters of a soil body to form a plurality of test working conditions, preparing a plurality of groups of soil body samples according to each working condition, and respectively carrying out multistage loading dynamic triaxial test on the soil body samples to generate hysteresis curves of various strain grades; Analyzing the condition that a hysteresis curve of a soil body sample at each strain level deviates from a self fitting curve and the condition that the hysteresis curve deviates from fitting curves of other soil body samples under the same working condition, obtaining a discrete coefficient of each soil body sample at each strain level, correcting an initial dynamic elastic modulus extracted from the hysteresis curve, determining a dynamic shear modulus and a damping ratio according to an elastic mechanics theory, fitting parameters of a nonlinear constitutive model, and establishing the nonlinear constitutive model responding to the change of the soil body state; Invoking a nonlinear constitutive model as material attribute in finite element simulation, changing soil physical parameters and compaction process parameters to simulate a vibration compaction process, analyzing vibration response characteristics to calculate compaction indexes representing compaction states of roadbed, combining the soil physical parameters and compaction process parameters, constructing feature vectors, applying simulated vehicle moving load to the roadbed subjected to compaction simulation, calculating rebound modulus of the top surface of the roadbed, and generating a simulation data set containing a plurality of feature vectors and rebound modulus thereof for training an integrated learning model; And according to the vibration response, compaction process parameters and soil physical parameters of the road roller in the construction site, acquiring the characteristic vector in real time, predicting the rebound modulus by using the integrated learning model, and evaluating the compaction quality of the roadbed.
2. The method for dynamically monitoring the rebound modulus of a roadbed according to claim 1, wherein the obtaining of the discrete coefficients of each soil sample at each strain level comprises: For each group of soil body samples under each test working condition, carrying out ellipse fitting on all data points distributed on hysteresis curves of all strain levels to obtain fitting ellipses, calculating fitting errors, taking the fitting ellipses as first characteristic values, calculating differences between the hysteresis curves of all strain levels and the fitting ellipses of other soil body samples under the same strain level as relative errors, taking the differences between the hysteresis curves of all strain levels and the fitting ellipses of other soil body samples as second characteristic values, respectively calculating correlation degrees of the first characteristic values and the second characteristic values of all strain levels between each group of soil body samples and the other soil body samples under each test working condition as first correlation degrees and second correlation degrees Guan Du, and carrying out positive mapping on average values of the first correlation degrees and the second correlation degrees as synergistic coefficients; aiming at each group of soil body samples under each test working condition, carrying out weighted fusion on the first characteristic value and the second characteristic value of each strain level to obtain weighted errors; the discrete coefficients are positively correlated with the weighted errors and negatively correlated with the co-coefficients.
3. A method of dynamically monitoring the rebound modulus of a roadbed as claimed in claim 1, wherein the correction of the initial dynamic elastic modulus extracted from the hysteresis curve comprises: aiming at hysteresis curves of each group of soil samples at each strain level under each test working condition, extracting the slope of the connecting line of two end points of each hysteresis curve as an initial dynamic elastic modulus; Then the first Under the test working condition Group soil body sample is at the first Dynamic elastic modulus after corresponding adjustment of each strain grade The calculation formula of (2) is as follows: , wherein, In order to achieve an initial modulus of elasticity in motion, Is the first Under the test working condition Group soil body sample is at the first The normalized discrete coefficients at the individual strain levels, The amplitude is adjusted for a preset.
4. A method for dynamically monitoring the rebound modulus of a roadbed according to claim 3, wherein the determining the dynamic shear modulus to damping ratio, fitting the parameters of the nonlinear constitutive model, and establishing the nonlinear constitutive model in response to the change of the soil state comprises: Calculating a dynamic shear modulus ratio and a damping ratio according to an elastic mechanics theory by utilizing the dynamic elastic modulus adjusted under each strain level for each group of soil body samples under each test working condition, and taking the average value of the dynamic shear modulus and the average value of the damping ratio of all soil body samples under the same strain level under each test working condition as the dynamic shear modulus and the damping ratio of each strain level under each test working condition; The method comprises the steps of converting axial dynamic strain corresponding to each strain level into shear strain, forming a three-dimensional array by the shear strain, dynamic shear modulus and damping ratio of each strain level, substituting Davidenkov the three-dimensional array of all strain levels under each test working condition into a model to perform fitting to obtain all model parameters, and constructing a prediction model of each model parameter relative to the physical parameters and mechanical environment parameters of a soil body by taking the model parameters and the mechanical environment parameters as independent variables aiming at the physical parameters and the mechanical environment parameters of the soil body under all test working conditions, so as to establish a Davidenkov model associated with the soil body state as a nonlinear constitutive model.
5. The dynamic monitoring method of the roadbed rebound modulus according to claim 1, wherein the construction process of the feature vector is as follows: Extracting vertical acceleration of the center of the vibration wheel at each moment when the vibration compaction process is simulated each time so as to calculate compaction indexes, wherein the compaction indexes comprise compaction metering values and compaction modulus; Calculating the ratio of energy corresponding to second harmonic in the spectrogram to energy corresponding to fundamental frequency, and taking the product of the energy corresponding to second harmonic and the energy corresponding to fundamental frequency as compaction value; Calculating dynamic contact force between vibrating wheel and roadbed soil The specific formula is as follows: , wherein, For the mass of the vibrating wheel, Is the vertical acceleration of the vehicle, and is the vertical acceleration, Indicating the magnitude of the exciting force, The angular frequency of the excitation force is indicated, As a cosine function, t represents a time variable, and the compaction modulus is iteratively solved by utilizing Lundberg contact theory; And forming a characteristic vector by using the physical parameters of the soil body, the compaction process parameters and the compaction indexes of each vibration compaction simulation.
6. The method of dynamic monitoring of the modulus of resilience of a subgrade according to claim 5, wherein said calculating the modulus of resilience of the subgrade top surface and generating a simulated data set comprising a plurality of eigenvectors and their modulus of resilience comprises: And extracting the maximum dynamic deflection value of the top surface of the roadbed by simulating the action of the moving load of the vehicle aiming at the roadbed model after each vibration compaction simulation, calculating the rebound modulus of the top surface of the roadbed according to an equivalent deflection method, and forming a simulation data set by using the feature vector obtained by multiple simulation and the corresponding rebound modulus.
7. The method for dynamically monitoring the rebound modulus of a roadbed according to claim 1, wherein the training process of the ensemble learning model is as follows: aiming at the simulation data set, taking the feature vectors and the rebound moduli of all samples as the input of a Relief-F algorithm, and calculating the important weights of the feature components on the rebound moduli; The method comprises the steps of calculating the correlation degree between any two feature components in feature vectors of all samples in a simulation data set, eliminating feature components with minimum important weights from any two feature components if the correlation degree is larger than a preset judging threshold value, otherwise, reserving the two feature components, defining the screened feature components as key components, and training an integrated learning model by taking the key components and rebound modulus in the feature vectors of all samples in the simulation data set as input.
8. The method for dynamically monitoring the rebound modulus of a roadbed according to claim 7, wherein the method for dynamically monitoring the rebound modulus of the roadbed is characterized in that the method comprises the steps of generating a rebound modulus area distribution cloud chart covering the whole roadbed by utilizing a trained integrated learning model according to the rebound modulus based on key components in a feature vector acquired in real time in a compacting operation, and compacting the rebound modulus at each position in the cloud chart if the rebound modulus is larger than or equal to a preset threshold value, otherwise, compacting the roadbed at the position is qualified, and otherwise, the position is not compacted.
9. A dynamic monitoring system for the rebound modulus of a subgrade, characterized in that the system comprises a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that the processor implements the steps of a dynamic monitoring method for the rebound modulus of a subgrade according to any of the claims 1-8 when executing the computer program.
10. A dynamic monitoring device for the resilient modulus of a subgrade, said device having stored therein a computer program, characterized in that said computer program, when executed by a processor, implements the steps of a dynamic monitoring method for the resilient modulus of a subgrade according to any of claims 1-8.

Description

Dynamic monitoring method, system and equipment for roadbed rebound modulus Technical Field The application relates to the technical field of rebound modulus monitoring, in particular to a dynamic monitoring method, a dynamic monitoring system and dynamic monitoring equipment for roadbed rebound modulus. Background Roadbed compaction is a key procedure in road engineering, the quality of the roadbed compaction directly influences the long-term performance and service life of a road surface, and the core evaluation index of the roadbed compaction is the rebound modulus of the top surface of the roadbed. The rebound modulus is used for representing the compressive strength and the deformation recovery capability of the roadbed under the load action, and is a basic basis for pavement structural design. When the rebound modulus of the roadbed is monitored, the vibration response of the road roller is monitored mainly by an intelligent compaction technology, so that the compaction state is evaluated, but as the soil body of the roadbed has nonlinearity, the modulus and damping characteristics of the soil body can change obviously along with the strain level and the stress state, so that principle errors exist when the soil body state is reversely pushed from the vibration response, and secondly, the differential vibration response can be excited at different positions by the same rolling process under the influence of the local heterogeneity of the soil body, so that the nonlinearity and the local heterogeneity of the soil body cannot be fully considered in the traditional method, the local state change is insensitive, the monitoring of the rebound modulus of the roadbed is inaccurate, and the evaluation accuracy of the compaction quality of the roadbed is affected. Disclosure of Invention In order to solve the technical problems, a dynamic monitoring method, a dynamic monitoring system and dynamic monitoring equipment for the rebound modulus of a roadbed are provided, so that the existing problems are solved. The application provides a dynamic monitoring method, a system and equipment for roadbed rebound modulus, which comprise the following steps: In a first aspect, an embodiment of the present application provides a method for dynamically monitoring a resilient modulus of a roadbed, the method including the steps of: Controlling physical parameters and mechanical environment parameters of a soil body to form a plurality of test working conditions, preparing a plurality of groups of soil body samples according to each working condition, and respectively carrying out multistage loading dynamic triaxial test on the soil body samples to generate hysteresis curves of various strain grades; Analyzing the condition that a hysteresis curve of a soil body sample at each strain level deviates from a self fitting curve and the condition that the hysteresis curve deviates from fitting curves of other soil body samples under the same working condition, obtaining a discrete coefficient of each soil body sample at each strain level, correcting an initial dynamic elastic modulus extracted from the hysteresis curve, determining a dynamic shear modulus and a damping ratio according to an elastic mechanics theory, fitting parameters of a nonlinear constitutive model, and establishing the nonlinear constitutive model responding to the change of the soil body state; Invoking a nonlinear constitutive model as material attribute in finite element simulation, changing soil physical parameters and compaction process parameters to simulate a vibration compaction process, analyzing vibration response characteristics to calculate compaction indexes representing compaction states of roadbed, combining the soil physical parameters and compaction process parameters, constructing feature vectors, applying simulated vehicle moving load to the roadbed subjected to compaction simulation, calculating rebound modulus of the top surface of the roadbed, and generating a simulation data set containing a plurality of feature vectors and rebound modulus thereof for training an integrated learning model; And according to the vibration response, compaction process parameters and soil physical parameters of the road roller in the construction site, acquiring the characteristic vector in real time, predicting the rebound modulus by using the integrated learning model, and evaluating the compaction quality of the roadbed. Preferably, the obtaining the discrete coefficient of each soil body sample at each strain level includes: For each group of soil body samples under each test working condition, carrying out ellipse fitting on all data points distributed on hysteresis curves of all strain levels to obtain fitting ellipses, calculating fitting errors, taking the fitting ellipses as first characteristic values, calculating differences between the hysteresis curves of all strain levels and the fitting ellipses of other soil body samples under the same strai