CN-121978214-A - Method and system for detecting compressive strength of concrete for constructional engineering

CN121978214ACN 121978214 ACN121978214 ACN 121978214ACN-121978214-A

Abstract

The invention discloses a method and a system for detecting the compressive strength of concrete for constructional engineering, and relates to the technical field of detection of the concrete strength of constructional engineering. And performing cross-modal fusion processing on the data set to generate a concrete internal defect characteristic map and a material uniformity distribution spectrum. And calling a pre-training strength prediction model to analyze the feature map and the distribution spectrum to obtain a compressive strength predicted value and a structural weak area mark. And carrying out space weight correction on the predicted value according to the mark to generate a corrected compressive strength value. Based on the correction value and the distribution spectrum, a detection report and a maintenance proposal are output. The method improves the intensity detection precision and the defect identification comprehensiveness.

Inventors

QU JIANMIN
LIU FANGLIANG
CAO LIFANG
Yu Kongan
WU LIJUAN
YANG BIN
LUO JUN

Assignees

湖南博联检测集团有限责任公司

Dates

Publication Date: 20260505
Application Date: 20260319

Claims (10)

1. The method for detecting the compressive strength of the concrete for the constructional engineering is characterized by comprising the following steps of: Collecting a multidimensional physical field data set of a concrete member, wherein the multidimensional physical field data set comprises ultrasonic propagation data, an infrared thermal imaging sequence and a three-dimensional surface morphology point cloud; Performing cross-modal fusion processing on the multidimensional physical field data set to generate a concrete internal defect characteristic map and a material uniformity distribution spectrum; Invoking a pre-trained strength prediction model to analyze the concrete internal defect characteristic map and the material uniformity distribution spectrum, and generating a compressive strength predicted value and a structural weak area mark; according to the structural weak area mark, carrying out space weight correction on the compressive strength predicted value to generate a corrected compressive strength value; and generating a detection report and a maintenance proposal scheme based on the corrected compressive strength value and the material uniformity distribution spectrum.
2. The method for detecting the compressive strength of concrete for constructional engineering according to claim 1, wherein the cross-modal fusion processing is performed on the multidimensional physical field data set to generate a concrete internal defect feature map and a material uniformity distribution spectrum, and the method comprises the following steps: performing time-frequency analysis processing on the ultrasonic wave propagation data, extracting propagation speed distribution, attenuation coefficient distribution and reflected signal energy distribution of ultrasonic waves in the concrete, and generating three-dimensional ultrasonic wave characteristic volume data; Performing thermal flow field inversion calculation on the infrared thermal imaging sequence, reconstructing heat conduction parameter distribution in the concrete, and generating three-dimensional thermodynamic characteristic volume data; carrying out structured light characteristic analysis on the three-dimensional surface morphology point cloud, calculating the distribution density and depth characteristics of surface microcracks, and generating three-dimensional morphology feature volume data; Establishing a unified space coordinate reference, and performing space registration and voxel alignment on the three-dimensional ultrasonic characteristic volume data, the three-dimensional thermodynamic characteristic volume data and the three-dimensional morphological characteristic volume data; Fusing a propagation speed value and an attenuation coefficient in the three-dimensional ultrasonic characteristic volume data, a heat conduction parameter value in the three-dimensional thermodynamic characteristic volume data and a microcrack depth characteristic in the three-dimensional morphological characteristic volume data at each aligned voxel position, and generating a concrete internal defect characteristic map through characteristic cascading and normalization; and calculating and generating the material uniformity distribution spectrum based on the uniformity of the reflected signal energy distribution in the three-dimensional ultrasonic characteristic volume data and the consistency of the heat conduction parameters in the three-dimensional thermodynamic characteristic volume data.
3. The method for detecting the compressive strength of concrete for constructional engineering according to claim 2, wherein the performing thermal field inversion calculation on the infrared thermal imaging sequence to reconstruct the thermal conduction parameter distribution inside the concrete and generate three-dimensional thermodynamic characteristic volume data comprises: Carrying out temperature field calibration and noise filtering on each frame of thermal image in the infrared thermal imaging sequence to obtain surface temperature field distribution data on a time sequence; establishing a three-dimensional heat conduction model inside the concrete according to a heat conduction equation, wherein the three-dimensional heat conduction model comprises initial estimated values of heat conductivity coefficient, specific heat capacity and density; Taking the surface temperature field distribution data on the time sequence as boundary conditions, and driving the three-dimensional heat conduction model to carry out iterative solution; in each iteration, adjusting the thermal conductivity parameter values of voxels within the three-dimensional thermal conductivity model such that the model models the differences between the calculated surface temperature field and the measured surface temperature field distribution data over the time series; And stopping iteration when the difference converges to a preset threshold value, wherein the heat conduction parameter value of each voxel in the three-dimensional heat conduction model forms the three-dimensional thermodynamic characteristic volume data.
4. The method for detecting the compressive strength of concrete for constructional engineering according to claim 1, wherein the step of calling a pre-trained strength prediction model to analyze the internal defect feature map of the concrete and the material uniformity distribution spectrum to generate a compressive strength predicted value and a structural weakness region mark comprises the following steps: inputting the internal defect feature map of the concrete into a defect feature extraction branch of the strength prediction model, wherein the defect feature extraction branch is constructed based on a convolutional neural network and is used for identifying and quantifying the size, shape and spatial position information of internal holes, cracks and segregation defects; Inputting the material uniformity distribution spectrum into a uniformity analysis branch of the intensity prediction model, wherein the uniformity analysis branch calculates a variation coefficient and a gradient variation trend of material properties in space; Fusing the quantized defect information output by the defect feature extraction branch and the material variation information output by the uniformity analysis branch, and calculating the intensity contribution value of each local area through a fully connected network layer; according to the intensity contribution values of all the local areas, the compressive strength predicted value is obtained through space integral calculation; And identifying a local area with the intensity contribution value lower than the global average contribution value by a certain proportion, extracting boundary coordinates of the local area, and generating the structural weak area mark.
5. The method for detecting the compressive strength of concrete for construction engineering according to claim 1, wherein the performing spatial weight correction on the predicted compressive strength value according to the structural weakness region mark to generate a corrected compressive strength value comprises: Determining the proportion of the weak area in the whole volume of the concrete member and the space distribution form according to the structural weak area mark; inquiring a preset weak area influence coefficient table, wherein the weak area influence coefficient table defines correction weights of weak areas with different duty ratios and different distribution forms for resisting the predicted value of the intensity of pressure; acquiring a correction weight coefficient matched with the current weak area occupation ratio and the distribution form according to the weak area influence coefficient table; Multiplying the compressive strength predicted value by the correction weight coefficient to obtain a preliminary correction strength value; And meanwhile, analyzing the overlapping degree of the structural weak area mark and the low-uniformity area in the material uniformity distribution spectrum, and performing secondary fine adjustment on the primary correction intensity value according to the overlapping degree to generate the corrected compressive intensity value.
6. The method for detecting the compressive strength of concrete for constructional engineering according to claim 4, wherein the identifying the local area with the strength contribution value lower than the global average contribution value by a certain proportion extracts the boundary coordinates thereof to generate the structural weakness area mark comprises the following steps: Calculating the average value of all local area intensity contribution values output by the intensity prediction model, and taking the average value as the global average contribution value; Setting a proportion threshold value, and identifying all local areas with intensity contribution values lower than the global average contribution value multiplied by the proportion threshold value; performing spatial cluster analysis on all the identified low-intensity contribution local areas, and merging the areas which are spatially adjacent and have similar attributes into the same weak area; extracting a voxel coordinate set of an outer contour of each combined weak area, and calculating the geometric center coordinate, equivalent diameter and volume of the weak area; and carrying out structural encapsulation on the voxel coordinate set, the geometric center coordinates, the equivalent diameter and the volume information of each weak area to form the structural weak area mark.
7. The method for detecting the compressive strength of concrete for construction engineering according to claim 1, wherein the generating a detection report and maintenance proposal based on the corrected compressive strength value and the material uniformity distribution spectrum comprises: Comparing the corrected compressive strength value with a design strength standard value to determine a strength grade assessment result; Analyzing the distribution range and the severity of a low-uniformity region in the uniformity distribution spectrum of the material, and evaluating the construction quality grade of the material; generating comprehensive evaluation of structural safety according to the strength grade evaluation result and the material construction quality grade; planning a specific repair reinforcing area according to the spatial position and the size information of the structural weak area mark; combining the comprehensive evaluation of the structural safety with the repair and reinforcement area, and matching corresponding construction process, material requirement and construction period estimation from a maintenance strategy library to form the maintenance proposal scheme; And integrating the strength grade assessment result, the material construction quality grade, the structural safety comprehensive evaluation and maintenance proposal scheme into the detection report in a standardized format.
8. A method for detecting the compressive strength of concrete for construction engineering according to claim 3, wherein the building of a three-dimensional heat conduction model inside the concrete according to a heat conduction equation comprises: Spatially discretizing the concrete member into a three-dimensional grid model consisting of regular hexahedral voxels; Giving an initial heat conductivity value, a specific heat capacity value and a density value to each voxel, wherein the initial value is determined based on the design proportion of the concrete and a standard material parameter library; Establishing a heat conduction relation between each voxel and adjacent voxels, and constructing a heat conduction equation set of the whole three-dimensional grid model based on a Fourier law and an energy conservation law; The heat conduction equation set takes the temperature change of each voxel along with time as an unknown quantity and takes the heat conduction coefficient, specific heat capacity and density among the voxels as parameters; The initial temperature field distribution of the model and external environmental boundary conditions are set, wherein the external environmental boundary conditions comprise an air convection heat exchange coefficient and solar radiation heat flow density.
9. The method for detecting the compressive strength of concrete for construction engineering according to claim 2, wherein the establishing a unified spatial coordinate reference, performing spatial registration and voxel alignment on the three-dimensional ultrasonic feature volume data, the three-dimensional thermodynamic feature volume data and the three-dimensional morphological feature volume data, comprises: Selecting at least three non-collinear identifiable common feature points from each data set, wherein the common feature points comprise component corner points, embedded mark points or obvious surface features; Calculating a space transformation matrix for mapping different data sets to the same coordinate system based on the common feature points, wherein the space transformation matrix comprises a rotation matrix and a translation vector; the space transformation matrix is applied to respectively conduct coordinate transformation on the three-dimensional ultrasonic characteristic volume data, the three-dimensional thermodynamic characteristic volume data and the three-dimensional morphological characteristic volume data; Defining a common three-dimensional grid containing the whole concrete member under the transformed unified coordinate system, wherein the voxel size of the grid is set to be not more than the resolution of the minimum data set; Resampling the parameter values in each data set to each voxel center point of the public three-dimensional grid through a cubic spline interpolation method to finish voxel alignment.
10. A concrete compressive strength detection system for construction engineering, characterized by comprising a processor and a memory, the memory storing a computer program, the processor implementing a concrete compressive strength detection method for construction engineering according to any one of claims 1 to 9 when executing the computer program.

Description

Method and system for detecting compressive strength of concrete for constructional engineering Technical Field The invention belongs to the technical field of detection of concrete strength of constructional engineering, and particularly relates to a method and a system for detecting concrete compressive strength of constructional engineering. Background Traditional concrete compressive strength detection mainly depends on a single-point coring test or a single physical field nondestructive detection method. Coring destroys structural integrity and sample representativeness is limited. Conventional nondestructive testing techniques such as ultrasonic, rebound or infrared thermography are often used independently. The ultrasonic wave propagation speed can indirectly reflect the intensity and the internal compactness, but the defect forms and the spatial distribution such as looseness, cracks and the like are difficult to accurately position. Infrared thermography techniques can capture surface temperature field differences to identify shallow void or water containing areas, but do not provide adequate quantitative assessment of deep defects and material uniformity. Three-dimensional laser scanning can acquire surface morphology, but cannot sense internal conditions. The isolated technical means provide single information dimension, lack of correlation fusion between data, and result in fuzzy quantitative description of internal defects of a concrete member, and inaccurate assessment of material spatial non-uniformity and strength dispersion caused by the material spatial non-uniformity. The existing strength prediction model is mostly based on single parameter regression, and the synergistic weakening effect of defect morphology, size, position and material non-uniformity on the whole bearing capacity is not comprehensively considered. Therefore, an evaluation method capable of integrating multidimensional information, realizing accurate quantitative characterization of defects and dynamically correlating the spatial distribution characteristics of defects with macroscopic intensity prediction is needed. Disclosure of Invention The present invention aims to solve at least one of the technical problems existing in the prior art; therefore, the invention provides a method and a system for detecting the compressive strength of concrete for constructional engineering, comprising the following steps: Collecting a multidimensional physical field data set of a concrete member, wherein the multidimensional physical field data set comprises ultrasonic propagation data, an infrared thermal imaging sequence and a three-dimensional surface morphology point cloud; Performing cross-modal fusion processing on the multidimensional physical field data set to generate a concrete internal defect characteristic map and a material uniformity distribution spectrum; Invoking a pre-trained strength prediction model to analyze the concrete internal defect characteristic map and the material uniformity distribution spectrum, and generating a compressive strength predicted value and a structural weak area mark; according to the structural weak area mark, carrying out space weight correction on the compressive strength predicted value to generate a corrected compressive strength value; and generating a detection report and a maintenance proposal scheme based on the corrected compressive strength value and the material uniformity distribution spectrum. Further, the cross-modal fusion processing is performed on the multidimensional physical field data set to generate a concrete internal defect feature map and a material uniformity distribution spectrum, including: performing time-frequency analysis processing on the ultrasonic wave propagation data, extracting propagation speed distribution, attenuation coefficient distribution and reflected signal energy distribution of ultrasonic waves in the concrete, and generating three-dimensional ultrasonic wave characteristic volume data; Performing thermal flow field inversion calculation on the infrared thermal imaging sequence, reconstructing heat conduction parameter distribution in the concrete, and generating three-dimensional thermodynamic characteristic volume data; carrying out structured light characteristic analysis on the three-dimensional surface morphology point cloud, calculating the distribution density and depth characteristics of surface microcracks, and generating three-dimensional morphology feature volume data; Establishing a unified space coordinate reference, and performing space registration and voxel alignment on the three-dimensional ultrasonic characteristic volume data, the three-dimensional thermodynamic characteristic volume data and the three-dimensional morphological characteristic volume data; Fusing a propagation speed value and an attenuation coefficient in the three-dimensional ultrasonic characteristic volume data, a heat conduction parameter value in the thr