A method and system for testing the compressive strength of concrete in building engineering.

Abstract

This invention discloses a method and system for testing the compressive strength of concrete in building engineering, relating to the field of concrete strength testing technology. The method includes: acquiring ultrasonic propagation data, infrared thermal imaging sequences, and three-dimensional surface topography point clouds of concrete components to form a multi-dimensional physical field data set; performing cross-modal fusion processing on the data set to generate a feature mapping map of internal defects in the concrete and a material homogeneity distribution spectrum; calling a pre-trained strength prediction model to analyze the feature mapping map and distribution spectrum to obtain predicted compressive strength values ​​and markers of structurally weak areas; spatially weighting the predicted values ​​based on the markers to generate corrected compressive strength values; and outputting a test report and maintenance recommendations based on the corrected values ​​and distribution spectrum. This method improves the accuracy of strength testing and the comprehensiveness of defect identification.

Description

Technical Field

[0001] This invention belongs to the field of concrete strength testing technology in building engineering, specifically a method and system for testing the compressive strength of concrete in building engineering. Background Technology

[0002] Traditional concrete compressive strength testing primarily relies on single-point core sampling or single-physical-field non-destructive testing methods. Core sampling damages structural integrity and has limited sample representativeness. Conventional non-destructive testing techniques, such as ultrasonic testing, rebound testing, or infrared thermography, are mostly applied independently. Ultrasonic propagation speed can indirectly reflect strength and internal density, but it is difficult to accurately locate the morphology and spatial distribution of defects such as porosity and cracks. Infrared thermography can capture surface temperature field differences to identify shallow voids or water-bearing areas, but its ability to quantitatively assess deep defects and material homogeneity is insufficient. Three-dimensional laser scanning can acquire surface morphology but cannot perceive internal conditions. These isolated techniques provide information with limited dimensions and lack correlation and integration between data, resulting in vague quantitative descriptions of internal defects in concrete components and inaccurate assessments of material spatial inhomogeneity and the resulting strength dispersion. Existing strength prediction models are mostly based on single-parameter regression and do not comprehensively consider the synergistic weakening effect of defect morphology, size, location, and material inhomogeneity on overall load-bearing capacity. Therefore, there is an urgent need for an assessment method that can integrate multi-dimensional information, achieve accurate quantitative characterization of defects, and dynamically correlate the spatial distribution characteristics of defects with macroscopic intensity prediction. Summary of the Invention

[0003] This invention aims to solve at least one of the technical problems existing in the prior art;

[0004] Therefore, this invention proposes a method and system for testing the compressive strength of concrete in building engineering, comprising:

[0005] A multi-dimensional physical field data set of concrete components is collected, which includes ultrasonic propagation data, infrared thermal imaging sequences, and three-dimensional surface topography point clouds.

[0006] The multi-dimensional physical field data set is subjected to cross-modal fusion processing to generate a feature mapping map of internal defects in concrete and a material homogeneity distribution spectrum.

[0007] The pre-trained strength prediction model is invoked to analyze the internal defect feature mapping map and material homogeneity distribution spectrum of the concrete, and to generate the predicted compressive strength value and the marker of the weak area of ​​the structure.

[0008] Based on the marking of the structural weak areas, the predicted compressive strength value is spatially weighted to generate the corrected compressive strength value.

[0009] Based on the corrected compressive strength value and material uniformity distribution spectrum, a test report and maintenance recommendations are generated.

[0010] Furthermore, the step of performing cross-modal fusion processing on the multi-dimensional physical field data set to generate a feature mapping map of internal defects in concrete and a material homogeneity distribution spectrum includes:

[0011] The ultrasonic propagation data is processed by time-frequency analysis to extract the propagation velocity distribution, attenuation coefficient distribution and reflected signal energy distribution of ultrasonic waves inside the concrete, and to generate three-dimensional ultrasonic feature data.

[0012] The infrared thermal imaging sequence is used to perform thermal flow field inversion calculations to reconstruct the distribution of heat conduction parameters inside the concrete and generate three-dimensional thermodynamic feature data.

[0013] Structured light feature analysis is performed on the three-dimensional surface topography point cloud to calculate the distribution density and depth features of surface microcracks and generate three-dimensional topography feature data.

[0014] A unified spatial coordinate reference is established, and the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data are spatially registered and voxel aligned.

[0015] At each aligned voxel position, the propagation velocity and attenuation coefficient from the three-dimensional ultrasonic feature data, the heat conduction parameter value from the three-dimensional thermodynamic feature data, and the microcrack depth feature from the three-dimensional morphology feature data are fused together. Through feature cascading and normalization, the feature mapping map of the internal defects of the concrete is generated.

[0016] Based on the consistency between the uniformity of reflected signal energy distribution in the three-dimensional ultrasonic feature data and the thermal conductivity parameters in the three-dimensional thermodynamic feature data, the uniformity distribution spectrum of the material is calculated and generated.

[0017] Further, the step of performing thermal flow field inversion calculation on the infrared thermal imaging sequence to reconstruct the distribution of heat conduction parameters inside the concrete and generate three-dimensional thermodynamic feature data includes:

[0018] Temperature field calibration and noise filtering are performed on each frame of the infrared thermal imaging sequence to obtain surface temperature field distribution data in the time series.

[0019] A three-dimensional heat conduction model of the interior of concrete is established based on the heat conduction equation. The three-dimensional heat conduction model includes initial estimates of thermal conductivity, specific heat capacity and density.

[0020] The surface temperature field distribution data over the time series is used as boundary conditions to drive the three-dimensional heat conduction model to perform iterative solutions.

[0021] In each iteration, the thermal conductivity parameter values ​​of the voxels inside the three-dimensional thermal conductivity model are adjusted to minimize the difference between the surface temperature field calculated by the model simulation and the measured surface temperature field distribution data on the time series.

[0022] When the difference converges to a preset threshold, the iteration stops. At this point, the thermal conductivity parameter value of each voxel in the three-dimensional thermal conductivity model constitutes the three-dimensional thermodynamic feature data.

[0023] Furthermore, the step of calling the pre-trained strength prediction model to analyze the feature mapping map of internal defects in the concrete and the material homogeneity distribution spectrum, generating predicted compressive strength values ​​and markers of structurally weak areas, includes:

[0024] The feature map of internal defects in the concrete is input into the defect feature extraction branch of the strength prediction model. The defect feature extraction branch is constructed based on a convolutional neural network and is used to identify and quantify the size, shape and spatial location information of internal pores, cracks and segregation defects.

[0025] The material uniformity distribution spectrum is input into the uniformity analysis branch of the strength prediction model, and the uniformity analysis branch calculates the coefficient of variation and gradient trend of the material properties in space.

[0026] The quantitative defect information output from the defect feature extraction branch and the material variation information output from the uniformity analysis branch are combined, and the intensity contribution value of each local region is calculated through a fully connected network layer.

[0027] The predicted compressive strength is calculated by spatial integration based on the strength contribution values ​​of all local regions.

[0028] Identify local areas where the intensity contribution value is lower than the global average contribution value by a certain proportion, extract their boundary coordinates, and generate the weak area markers of the structure.

[0029] Further, the step of spatially weighting the predicted compressive strength value based on the markers of the structurally weak areas to generate a corrected compressive strength value includes:

[0030] Based on the marked weak areas of the structure, determine the proportion and spatial distribution of the weak areas in the overall volume of the concrete component;

[0031] Query the preset weak area influence coefficient table, which defines the correction weights for the predicted compressive strength values ​​of weak areas with different proportions and distribution patterns.

[0032] Based on the table of influence coefficients for weak areas, obtain the corrected weighting coefficients that match the current proportion and distribution pattern of weak areas;

[0033] Multiply the predicted compressive strength value by the correction weighting coefficient to obtain the preliminary corrected strength value;

[0034] Simultaneously, the degree of overlap between the structural weak area markers and the low uniformity areas in the material uniformity distribution spectrum is analyzed, and the preliminary corrected strength value is finely adjusted based on the degree of overlap to generate the corrected compressive strength value.

[0035] Furthermore, for local regions where the identification intensity contribution value is lower than the global average contribution value by a certain proportion, their boundary coordinates are extracted to generate the structurally weak region markers, including:

[0036] The average value of the intensity contribution values ​​of all local regions output by the intensity prediction model is calculated as the global average contribution value.

[0037] A proportional threshold is set to identify all local regions whose intensity contribution value is lower than the global average contribution value multiplied by the proportional threshold.

[0038] Spatial clustering analysis was performed on all identified low-intensity contributing local areas to merge spatially adjacent and similar-attributed areas into the same weak area.

[0039] For each merged weak region, extract the set of voxel coordinates of its outer contour, and calculate the geometric center coordinates, equivalent diameter, and volume of the weak region;

[0040] The set of voxel coordinates, geometric center coordinates, equivalent diameter and volume information of each weak region are structurally encapsulated to form the weak region marker.

[0041] Furthermore, the step of generating a test report and maintenance recommendations based on the corrected compressive strength value and material uniformity distribution spectrum includes:

[0042] The corrected compressive strength value is compared with the design strength standard value to determine the strength grade evaluation result;

[0043] Analyze the distribution range and severity of low uniformity regions in the material uniformity distribution spectrum to assess the material construction quality grade;

[0044] Based on the strength grade assessment results and the material construction quality grade, a comprehensive structural safety evaluation is generated.

[0045] Based on the spatial location and size information of the marked weak areas in the structure, specific repair and reinforcement areas are planned;

[0046] Based on the comprehensive evaluation of structural safety and the repair and reinforcement areas, the corresponding construction techniques, material requirements and construction period estimates are matched from the maintenance strategy library to form the maintenance recommendation plan;

[0047] The strength grade assessment results, material construction quality grade, comprehensive evaluation of structural safety, and maintenance recommendations are integrated into a standardized test report.

[0048] Furthermore, the establishment of a three-dimensional heat conduction model inside concrete based on the heat conduction equation includes:

[0049] The concrete component is discretized in space into a three-dimensional mesh model composed of regular hexahedral voxels;

[0050] Each voxel is assigned an initial value for thermal conductivity, specific heat capacity, and density. The initial values ​​are determined based on the concrete mix design and a standard material parameter library.

[0051] Establish the heat conduction relationship between each voxel and its neighboring voxels, and construct the heat conduction equation set for the entire three-dimensional mesh model based on Fourier's law and the law of conservation of energy.

[0052] The heat conduction equations take the temperature change of each voxel over time as the unknown quantity and the thermal conductivity, specific heat capacity and density between voxels as parameters.

[0053] The initial temperature field distribution and external environmental boundary conditions of the model are set, including the air convection heat transfer coefficient and the solar radiation heat flux density.

[0054] Furthermore, the establishment of a unified spatial coordinate reference, and the spatial registration and voxel alignment of the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data, include:

[0055] In each dataset, select at least three non-collinear, identifiable common feature points, including component corner points, pre-embedded marker points, or significant surface features;

[0056] Based on the common feature points, a spatial transformation matrix is ​​calculated to map different data sets to the same coordinate system. The spatial transformation matrix includes a rotation matrix and a translation vector.

[0057] The spatial transformation matrix is ​​used to perform coordinate transformations on the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data, respectively.

[0058] Under the transformed unified coordinate system, a common three-dimensional mesh containing the entire concrete component is defined, and the voxel size of the mesh is set to be no greater than the resolution of the minimum data set;

[0059] The parameter values ​​in each dataset are resampled to the center point of each voxel in the common 3D mesh using cubic spline interpolation to complete voxel alignment.

[0060] Furthermore, the present invention also includes a concrete compressive strength testing system for building engineering, the system including a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the above-described concrete compressive strength testing method for building engineering.

[0061] Compared with the prior art, the beneficial effects of the present invention are:

[0062] A multi-dimensional physical field data set is constructed by collecting ultrasonic propagation data, infrared thermal imaging sequences, and three-dimensional surface topography point clouds. Ultrasonic data contains information on internal density and defects, infrared sequences reflect near-surface thermal property differences, and surface point clouds record macroscopic geometric deformations. Cross-modal fusion processing of these three heterogeneous data types is not a simple superposition, but rather a process involving feature alignment, information complementarity, and redundancy verification to construct a concrete internal defect feature map and a material homogeneity distribution spectrum that simultaneously express the spatial morphology of internal defects and the continuous changes in material properties. This process achieves comprehensive digital analysis of concrete components from the interior to the surface, and from microscopic defects to macroscopic property distribution, breaking through the information barriers of single detection technologies. The defect feature map clearly presents the location, shape, and size of cracks, pores, etc., while the homogeneity distribution spectrum quantifies the gradient changes in material properties. This provides a high-fidelity comprehensive state characterization that is unattainable by traditional methods for subsequent precise analysis.

[0063] The pre-trained strength prediction model is used to analyze the generated defect feature map and uniformity distribution spectrum. Trained with massive amounts of fused data, this model understands the morphological characteristics, density, and mechanical significance of various defects within a uniform context. The model not only outputs the overall compressive strength prediction but, more importantly, automatically identifies and marks structurally weak areas that significantly impact load-bearing capacity. Based on this, the initial overall strength prediction is spatially weighted and corrected according to the spatial location, extent, and severity of these marked weak areas. This correction process simulates the stress concentration and load-bearing capacity reduction effects in defect areas under actual stress conditions, dynamically transforming the spatial distribution information of defects into a correction to the macroscopic strength value. This ensures that the final corrected compressive strength value is no longer a general average estimate but a more accurate assessment of the component's actual stress performance, fully considering the adverse effects of spatial defect distribution, significantly improving the reliability of the prediction and its engineering guidance value. Attached Figure Description

[0064] Figure 1This is a flowchart illustrating the steps of the concrete compressive strength testing method and system for building engineering described in this invention.

[0065] Figure 2 The curve showing the convergence verification of the inversion and iteration of the thermal flow field of concrete.

[0066] Figure 3 The flowchart for the intensity prediction model analysis;

[0067] Figure 4 A composite graph of bar and line graphs showing the multi-dimensional evaluation scores of concrete structures.

[0068] Figure 5 A bar chart comparing the corrected compressive strength of concrete components. Detailed Implementation

[0069] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] Please see Figure 1 This invention relates to a method for testing the compressive strength of concrete in building engineering. This method achieves accurate prediction of compressive strength and structural assessment through the acquisition and fusion of multi-dimensional physical field data sets. The method includes the following steps: acquiring a multi-dimensional physical field data set of concrete components, including ultrasonic wave propagation data, infrared thermal imaging sequences, and three-dimensional surface topography point clouds; performing cross-modal fusion processing on the multi-dimensional physical field data set to generate a feature mapping map of internal defects in the concrete and a material homogeneity distribution spectrum; calling a pre-trained strength prediction model to analyze the feature mapping map of internal defects in the concrete and the material homogeneity distribution spectrum to generate predicted compressive strength values ​​and structural weak area markers; based on the structural weak area markers, spatial weight correction is applied to the predicted compressive strength values ​​to generate corrected compressive strength values; based on the corrected compressive strength values ​​and the material homogeneity distribution spectrum, a test report and maintenance recommendations are generated.

[0071] In one embodiment of the present invention, generating a feature map of internal defects in concrete and a material homogeneity distribution spectrum in cross-modal fusion processing includes multiple steps. Time-frequency analysis is performed on ultrasonic propagation data to extract the propagation velocity distribution, attenuation coefficient distribution, and reflected signal energy distribution of ultrasonic waves within the concrete, generating three-dimensional ultrasonic feature volume data. Thermal flow field inversion calculations are performed on infrared thermal imaging sequences to reconstruct the distribution of heat conduction parameters within the concrete, generating three-dimensional thermodynamic feature volume data. Structured light feature analysis is performed on the three-dimensional surface morphology point cloud to calculate the distribution density and depth characteristics of surface microcracks, generating three-dimensional morphology feature volume data. A unified spatial coordinate benchmark is established, and the three-dimensional ultrasonic feature data, three-dimensional thermodynamic feature data, and three-dimensional morphological feature data are spatially registered and voxel aligned. At each aligned voxel position, the propagation velocity and attenuation coefficient values ​​from the three-dimensional ultrasonic feature data, the thermal conductivity parameter values ​​from the three-dimensional thermodynamic feature data, and the microcrack depth features from the three-dimensional morphological feature data are integrated. Through feature cascading and normalization, a feature mapping map of internal defects in concrete is generated. Based on the consistency between the uniformity of reflected signal energy distribution in the three-dimensional ultrasonic feature data and the thermal conductivity parameter in the three-dimensional thermodynamic feature data, a material homogeneity distribution spectrum is calculated and generated. The specific process of spatial registration and voxel alignment is as follows: At least three non-collinear, identifiable common feature points are selected in each dataset. These common feature points include component corner points, embedded marker points, or significant surface features. Based on these common feature points, a spatial transformation matrix is ​​calculated to map different datasets to the same coordinate system. This spatial transformation matrix includes a rotation matrix and a translation vector. The spatial transformation matrix is ​​then applied to transform the coordinates of the 3D ultrasonic feature data, 3D thermodynamic feature data, and 3D morphological feature data, respectively. Under the transformed unified coordinate system, a common 3D mesh encompassing the entire concrete component is defined, with the voxel size set to be no larger than the resolution of the smallest dataset. The parameter values ​​in each dataset are resampled to the center point of each voxel in the common 3D mesh using cubic spline interpolation, thus completing the voxel alignment.

[0072] In practical implementation, time-frequency analysis is performed on ultrasonic propagation data to extract the propagation velocity distribution, attenuation coefficient distribution, and reflected signal energy distribution of ultrasonic waves within concrete, generating three-dimensional ultrasonic feature volume data. Thermal flow field inversion calculations are performed on infrared thermal imaging sequences to reconstruct the distribution of heat conduction parameters within the concrete, generating three-dimensional thermodynamic feature volume data. Structured light feature analysis is performed on the three-dimensional surface morphology point cloud to calculate the distribution density and depth characteristics of surface microcracks, generating three-dimensional morphology feature volume data. A unified spatial coordinate benchmark is established, and the three-dimensional ultrasonic feature volume data, three-dimensional thermodynamic feature volume data, and three-dimensional morphology feature volume data are spatially registered and voxel aligned. At each aligned voxel position, the propagation velocity and attenuation coefficient values ​​from the three-dimensional ultrasonic feature volume data, the heat conduction parameter values ​​from the three-dimensional thermodynamic feature data, and the microcrack depth characteristics from the three-dimensional morphology feature data are fused. Through feature cascading and normalization, a feature mapping map of internal defects in concrete is generated. Based on the consistency between the uniformity of reflected signal energy distribution in the three-dimensional ultrasonic feature volume data and the heat conduction parameters in the three-dimensional thermodynamic feature data, a material homogeneity distribution spectrum is calculated and generated.

[0073] In some embodiments, the spatial registration and voxel alignment process involves selecting at least three non-collinear, identifiable common feature points. These common feature points include component corner points, embedded marker points, or significant surface features. Based on these common feature points, a spatial transformation matrix is ​​calculated to map different data sets to the same coordinate system. The spatial transformation matrix includes a rotation matrix and a translation vector. Applying the spatial transformation matrix, coordinate transformations are performed on the 3D ultrasonic feature data, 3D thermodynamic feature data, and 3D morphological feature data, respectively. Under the transformed unified coordinate system, a common 3D mesh encompassing the entire concrete component is defined. The voxel size of the mesh is set to be no greater than the resolution of the smallest data set. The parameter values ​​in each data set are resampled to the center point of each voxel in the common 3D mesh using cubic spline interpolation, completing the voxel alignment. In a specific implementation, the coordinate transformation is achieved using the following formula:

[0074] ;

[0075] in: Represents the three-dimensional coordinate vector of a point in the original dataset; Represents the rotation matrix; Represents the translation vector; This represents a three-dimensional coordinate vector in the unified coordinate system after transformation. It can be understood as a rotation matrix. It is a 3×3 orthogonal matrix, and the translation vector is... It is a 3×1 column vector, and the selection of common feature points ensures the accuracy of the spatial transformation matrix calculation.

[0076] Optionally, during feature concatenation and normalization, the feature vector at each voxel location is composed of the propagation velocity value, attenuation coefficient, thermal conductivity parameter value from the three-dimensional ultrasonic feature data, and microcrack depth feature from the three-dimensional morphology feature data, all concatenated. The normalization operation scales each feature dimension to a zero-mean, unit variance range. In some embodiments, the material homogeneity distribution spectrum is calculated based on the local variance of the reflected signal energy distribution in the three-dimensional ultrasonic feature data and the local gradient magnitude of the thermal conductivity parameter value in the three-dimensional thermodynamic feature data, generating a homogeneity index value for each voxel location through weighted summation. It can be understood that the uniformity of the reflected signal energy distribution is obtained by calculating the ratio of the standard deviation to the mean of the energy values ​​within a local region, and the consistency of the thermal conductivity parameter is obtained by calculating the coefficient of variation of the parameter values ​​within a local region. In specific implementations, the generation of the concrete internal defect feature map also involves convolutional neural network processing of the fused feature vectors to enhance the identification ability of defect boundaries, but feature concatenation and normalization are fundamental steps in generating the defect feature map.

[0077] In one embodiment of the present invention, the thermal flow field inversion calculation reconstructs the distribution of thermal conduction parameters inside concrete, generating three-dimensional thermodynamic feature data. Temperature field calibration and noise filtering are performed on each frame of the infrared thermal imaging sequence to obtain surface temperature field distribution data over time. A three-dimensional thermal conduction model of the concrete interior is established based on the thermal conduction equation. The three-dimensional thermal conduction model includes initial estimates of thermal conductivity, specific heat capacity, and density. The surface temperature field distribution data over time is used as boundary conditions to drive the iterative solution of the three-dimensional thermal conduction model. In each iteration, the thermal conduction parameter values ​​of the voxels within the three-dimensional thermal conduction model are adjusted to minimize the difference between the surface temperature field calculated by the model and the measured surface temperature field distribution data over time. When the difference converges to a preset threshold, the iteration stops. At this point, the thermal conduction parameter values ​​of each voxel in the three-dimensional thermal conduction model constitute the three-dimensional thermodynamic feature data. The specific method for establishing a three-dimensional heat conduction model is as follows: the concrete component is discretized in space into a three-dimensional mesh model composed of regular hexahedral voxels; each voxel is assigned an initial thermal conductivity value, specific heat capacity value, and density value, the initial values ​​being determined based on the concrete design mix proportion and standard material parameter library; the heat conduction relationship between each voxel and its adjacent voxels is established, and based on Fourier's law and the law of conservation of energy, a set of heat conduction equations for the entire three-dimensional mesh model is constructed; the heat conduction equations use the temperature change of each voxel over time as the unknown quantity, and the thermal conductivity, specific heat capacity, and density between voxels as parameters; the initial temperature field distribution and external environmental boundary conditions of the model are set, including the air convection heat transfer coefficient and the solar radiation heat flux density.

[0078] In practice, thermal flow field inversion calculations are performed on the infrared thermal imaging sequence to reconstruct the distribution of thermal conduction parameters inside the concrete and generate three-dimensional thermodynamic feature data. This process first requires temperature field calibration and noise filtering for each frame of the infrared thermal imaging sequence to obtain surface temperature field distribution data over time. A three-dimensional thermal conduction model of the concrete interior is established based on the thermal conduction equation. This model includes initial estimates of thermal conductivity, specific heat capacity, and density. The surface temperature field distribution data over time is used as boundary conditions to drive iterative solutions for the three-dimensional thermal conduction model. In each iteration, the thermal conduction parameter values ​​of the voxels within the three-dimensional thermal conduction model are adjusted to minimize the difference between the simulated surface temperature field and the measured surface temperature field distribution data over time. When the difference converges to a preset threshold, the iteration stops. At this point, the thermal conduction parameter values ​​of each voxel in the three-dimensional thermal conduction model constitute the three-dimensional thermodynamic feature data.

[0079] In some embodiments, the specific method for establishing a three-dimensional heat conduction model inside concrete is to discretize the concrete component in space into a three-dimensional mesh model composed of regular hexahedral voxels. Each voxel is assigned an initial thermal conductivity value, specific heat capacity value, and density value, which are determined based on the concrete mix design and a standard material parameter library. It can be understood that the initial thermal conductivity, specific heat capacity, and density values ​​of each voxel are set to be uniformly distributed or partitioned according to the mix design information. The heat conduction relationship between each voxel and its neighboring voxels is established, and based on Fourier's law and the law of conservation of energy, a set of heat conduction equations for the entire three-dimensional mesh model is constructed. In a specific implementation, the set of heat conduction equations for the three-dimensional mesh model can be expressed as:

[0080] ;

[0081] in: Material density representing voxels; The specific heat capacity of a voxel; Indicates the thermal conductivity of a voxel; The temperature of a voxel; Indicates time; It is a gradient operator. The heat conduction equations use the temperature change of each voxel over time as the unknown quantity, and the thermal conductivity, specific heat capacity and density between voxels as parameters to set the initial temperature field distribution and external environmental boundary conditions of the model. The external environmental boundary conditions include the air convection heat transfer coefficient and the solar radiation heat flux density.

[0082] Optionally, the iterative solution process employs gradient descent or conjugate gradient methods to minimize the difference between the simulated and measured surface temperature field distribution data. The difference function is typically defined as the sum of squares of the differences between the simulated and measured temperatures at all time steps. In some embodiments, a preset threshold is set such that the difference function value decreases by more than 99.5% relative to the initial iterative calculation value. In each iteration, the thermal conductivity parameters of the voxels within the three-dimensional thermal conductivity model are adjusted and updated according to the backpropagation gradient direction. It can be understood that adjusting the thermal conductivity parameters of the voxels within the three-dimensional thermal conductivity model only applies to the internal voxels; the thermal conductivity parameters of the boundary voxels remain unchanged due to constraints imposed by the measured data. When the iteration stops, the set of current thermal conductivity, specific heat capacity, and density values ​​of all voxels in the three-dimensional thermal conductivity model is output as three-dimensional thermodynamic feature data. In a specific implementation, the air convection heat transfer coefficient and solar radiation heat flux density are obtained from environmental monitoring data as known input parameters and are consistent with the actual conditions during infrared thermal imaging acquisition.

[0083] See Figure 2 This is a convergence verification curve for the concrete thermal flow field inversion iteration, reflecting the difference between the simulated surface temperature field and the measured temperature field during the iterative solution of the heat conduction model. This curve verifies the effectiveness of the thermal flow field inversion algorithm. By iteratively adjusting the heat conduction parameters of the voxels inside the concrete, the simulated temperature field ultimately achieves a high degree of consistency with the measured temperature field. When the curve converges, the output three-dimensional thermodynamic feature data can be directly used for subsequent concrete compressive strength testing and defect identification. The setting of the convergence iteration number and threshold also provides a reference for balancing the computational efficiency and accuracy of thermal flow field inversion in engineering. It directly proves the feasibility of the thermal flow field inversion algorithm; through iterative optimization, the model can converge to the preset accuracy, ensuring the reliability of the subsequent three-dimensional thermodynamic feature data.

[0084] See Figure 3In one embodiment of the present invention, the process of generating predicted compressive strength values ​​and structurally weak area markers through strength prediction model analysis involves multiple branches. The feature map of internal concrete defects is input into the defect feature extraction branch of the strength prediction model. This branch, constructed based on a convolutional neural network, is used to identify and quantify the size, shape, and spatial location information of internal pores, cracks, and segregation defects. The material homogeneity distribution spectrum is input into the homogeneity analysis branch of the strength prediction model. This branch calculates the coefficient of variation and gradient trend of material properties in space. The quantified defect information output from the defect feature extraction branch and the material variation information output from the homogeneity analysis branch are fused, and the strength contribution value of each local region is calculated through a fully connected network layer. Based on the strength contribution values ​​of all local regions, the predicted compressive strength value is calculated through spatial integration. Local regions whose strength contribution values ​​are lower than the global average contribution value by a certain proportion are identified, and their boundary coordinates are extracted to generate structurally weak area markers. The specific steps for generating structurally weak area markers are as follows: Calculate the average value of the intensity contribution of all local regions output by the intensity prediction model as the global average contribution value; set a proportional threshold to identify all local regions whose intensity contribution value is lower than the global average contribution value multiplied by the proportional threshold; perform spatial clustering analysis on all identified low-intensity contribution local regions, merging spatially adjacent regions with similar attributes into the same weak area; for each merged weak area, extract the voxel coordinate set of its outer contour, and calculate the geometric center coordinates, equivalent diameter, and volume of the weak area; encapsulate the voxel coordinate set, geometric center coordinates, equivalent diameter, and volume information of each weak area in a structured manner to form a structurally weak area marker.

[0085] In practical implementation, a pre-trained strength prediction model is invoked to analyze the feature map of internal defects in concrete and the material homogeneity distribution spectrum, generating predicted compressive strength values ​​and markings of structurally weak areas. The feature map of internal defects in concrete is input into the defect feature extraction branch of the strength prediction model. This branch, built on a convolutional neural network, is used to identify and quantify the size, shape, and spatial location information of internal pores, cracks, and segregation defects. The material homogeneity distribution spectrum is input into the homogeneity analysis branch of the strength prediction model, which calculates the coefficient of variation and gradient trend of material properties in space. The quantified defect information output from the defect feature extraction branch and the material variation information output from the homogeneity analysis branch are fused. The strength contribution value of each local region is calculated through a fully connected network layer. Based on the strength contribution values ​​of all local regions, the predicted compressive strength value is obtained through spatial integration. The spatial integration formula is expressed as:

[0086] ;

[0087] in: This represents the predicted compressive strength value; Represents spatial coordinates The intensity contribution function at that location; This represents the three-dimensional spatial integral domain corresponding to the concrete component. Local regions where the strength contribution value is lower than the global average contribution value by a certain proportion are identified, and their boundary coordinates are extracted to generate markers for structurally weak areas.

[0088] In some embodiments, the specific steps for generating structurally weak region markers include calculating the average of the intensity contribution values ​​of all local regions output by the intensity prediction model as the global average contribution value; setting a proportional threshold; and identifying all local regions whose intensity contribution value is lower than the global average contribution value multiplied by the proportional threshold. Spatial clustering analysis is performed on all identified low-intensity contribution local regions to merge spatially adjacent regions with similar attributes into the same weak region. The spatial clustering analysis uses a clustering algorithm based on Euclidean distance and intensity contribution value similarity. The proportional threshold is a preset constant used to control the sensitivity of the structurally weak region markers, and its value ranges from 0.5 to 0.8. For each merged weak region, the voxel coordinate set of its outer contour is extracted, and the geometric center coordinates, equivalent diameter, and volume of the weak region are calculated. The geometric center coordinates are obtained by calculating the arithmetic mean of all voxel coordinates within the weak region, and the equivalent diameter is defined as the diameter of a sphere with the same volume as the weak region. The voxel coordinate set, geometric center coordinates, equivalent diameter, and volume information of each weak region are structurally encapsulated to form the structurally weak region marker.

[0089] Optionally, the convolutional neural network for the defect feature extraction branch includes multiple convolutional and pooling layers to extract multi-level defect features from the internal defect feature map of concrete. The fully connected network layer receives the fused feature vector and outputs the strength contribution value for each local region. In some embodiments, the homogeneity analysis branch calculates the coefficient of variation of material properties in space using a local window statistical method, and the gradient change trend is calculated using the Sobel operator or a similar edge detection operator. It can be understood that a local region corresponds to a voxel in a three-dimensional spatial grid or a small block composed of multiple voxels, and the calculation of the strength contribution value depends on a linear or nonlinear combination of defect features and material homogeneity features. Optionally, spatial integration is approximated on a discrete voxel grid through summation; the strength contribution value of each voxel is multiplied by the voxel volume and then summed to obtain the predicted compressive strength value.

[0090] In one embodiment of the present invention, the spatial weight correction to generate the corrected compressive strength value includes the following operations: Based on the structural weak area markers, determine the proportion and spatial distribution of the weak areas in the overall volume of the concrete member; query a pre-set weak area influence coefficient table, which defines the correction weights for the predicted compressive strength value of weak areas with different proportions and distribution patterns; obtain correction weight coefficients that match the current weak area proportion and distribution pattern based on the weak area influence coefficient table; multiply the predicted compressive strength value by the correction weight coefficients to obtain the preliminary corrected strength value; simultaneously, analyze the degree of overlap between the structural weak area markers and the low-uniformity areas in the material homogeneity distribution spectrum, and perform a secondary fine-tuning of the preliminary corrected strength value based on the degree of overlap to generate the corrected compressive strength value.

[0091] In practice, the predicted compressive strength value is spatially weighted and corrected based on the weak area markers to generate a corrected compressive strength value. This process first requires determining the proportion and spatial distribution of weak areas within the overall volume of the concrete member based on the weak area markers. The weak area markers contain volume information and spatial coordinates for each weak area. The volume proportion is calculated by summing the volumes of all weak areas and comparing them with the total volume of the concrete member. The spatial distribution is classified by analyzing the relative positions and geometric characteristics between weak areas. A pre-set weak area influence coefficient table is consulted, which defines the correction weights for the predicted compressive strength value for weak areas with different proportions and distribution patterns. Correction weight coefficients matching the current weak area proportion and distribution pattern are obtained from this table. The predicted compressive strength value is multiplied by the correction weight coefficient to obtain the initial corrected strength value. Simultaneously, the degree of overlap between the weak area markers and low-uniformity areas in the material homogeneity distribution spectrum is analyzed. Based on the degree of overlap, the initial corrected strength value is fine-tuned to generate the corrected compressive strength value.

[0092] In some embodiments, the weak area influence coefficient table is stored in the form of a data table. The table rows correspond to different weak area volume percentage ranges, and the table columns correspond to different spatial distribution morphology types. Each cell stores the corresponding corrected weight coefficient value. The spatial distribution morphology type is classified based on the degree of clustering and shape regularity of the weak areas, including discrete distribution, banded distribution, and clustered distribution. Discrete distribution indicates that the weak areas are isolated and scattered in space; banded distribution indicates that the weak areas extend along a certain direction; and clustered distribution indicates that the weak areas are densely clustered in a local area. It can be understood that when determining the spatial distribution morphology of the weak areas, the clustering characteristics and contour length ratio of the geometric center coordinates of all weak areas are calculated. When the number of weak areas is less than three, it is directly classified as a discrete distribution. The corrected compressive strength value is calculated using the following formula:

[0093] ;

[0094] in: This indicates the corrected compressive strength value; This represents the predicted compressive strength value; This represents the corrected weighting coefficient obtained from the table of influence coefficients for weak areas; This indicates the preset fine-tuning coefficient; This indicates the degree of overlap between the markers indicating structurally weak areas and the low-uniformity areas in the material homogeneity distribution spectrum. (Overlap degree) Defined as the percentage of the spatial volume of the intersection between structurally weak regions and regions in the material uniformity distribution spectrum where the uniformity index is below a threshold.

[0095] Optionally, the content of the weak area influence coefficient table is generated based on historical experimental data and mechanical model simulation calibration. See Table 1 for a simplified weak area influence coefficient.

[0096] Table 1: Data Table of Impact Coefficients of Vulnerable Areas

[0097]

[0098] In some embodiments, a linear interpolation method is used to obtain the corrected weighting coefficients. When the actual volume percentage or distribution pattern of the weak area falls between the discrete values ​​defined in the table, interpolation is performed using adjacent data. It can be understood that the fine-tuning coefficients... It is a constant between zero and one, used to control the influence of the degree of overlap on the correction intensity value. The calculation is achieved through a three-dimensional spatial voxel traversal, counting the proportion of voxels that simultaneously belong to both structurally weak region markers and low-uniformity regions out of the total number of weak region voxels. Optionally, the secondary fine-tuning operation directly multiplies the initial corrected intensity value by an adjustment factor. Completed, the adjustment factor ensures that the corrected compressive strength value further decreases as the degree of overlap increases. In specific implementation, the low uniformity region in the material uniformity distribution spectrum is obtained by binarization segmentation by setting a uniformity index threshold, which is determined based on the standard uniformity distribution of concrete materials.

[0099] See Figure 4This is a composite bar and line graph comparing the scores of various core indicators in the concrete structure's compressive strength testing process. It visually demonstrates the gap between the actual performance and standard requirements of these indicators. All actual scores are higher than the corresponding standard values, indicating that the concrete component's strength, uniformity, structural integrity, and safety meet engineering requirements, and its overall performance is reliable. The actual score for "Material Uniformity" is closest to the standard value, representing a relatively weak link in the process; the impact of low uniformity areas on long-term structural stability needs attention. While the actual score for "Maintenance Priority" (approximately 75 points) is higher than the standard value (70 points), it still suggests the need to plan preventative maintenance measures. The graph visually presents the compliance level of each indicator, serving as visual support for the test report and enhancing the persuasiveness of the conclusions. It provides a basis for allocating maintenance resources, prioritizing weak areas related to material uniformity.

[0100] In one embodiment of the present invention, the generation of the inspection report and maintenance recommendation scheme is based on the corrected compressive strength value and the material homogeneity distribution spectrum. The corrected compressive strength value is compared with the design strength standard value to determine the strength grade assessment result; the distribution range and severity of low homogeneity areas in the material homogeneity distribution spectrum are analyzed to assess the material construction quality grade; based on the strength grade assessment result and the material construction quality grade, a comprehensive structural safety evaluation is generated; based on the spatial location and size information of the structural weak area markers, specific repair and reinforcement areas are planned; combining the comprehensive structural safety evaluation and the repair and reinforcement areas, corresponding construction techniques, material requirements, and construction period estimates are matched from the maintenance strategy library to form a maintenance recommendation scheme; the strength grade assessment result, material construction quality grade, comprehensive structural safety evaluation, and maintenance recommendation scheme are integrated into a standardized format inspection report.

[0101] In practice, a test report and maintenance recommendations are generated based on the corrected compressive strength value and the material uniformity distribution spectrum. The corrected compressive strength value is compared with the design strength standard value to determine the strength grade assessment result. The distribution range and severity of low uniformity areas in the material uniformity distribution spectrum are analyzed to assess the material construction quality grade. A comprehensive structural safety evaluation is generated based on the strength grade assessment result and the material construction quality grade. Specific repair and reinforcement areas are planned based on the spatial location and size information of the structural weak areas. Combining the comprehensive structural safety evaluation and the repair and reinforcement areas, corresponding construction techniques, material requirements, and schedule estimates are matched from the maintenance strategy library to form a maintenance recommendation plan. The strength grade assessment result, material construction quality grade, comprehensive structural safety evaluation, and maintenance recommendation plan are integrated into a standardized test report.

[0102] In some embodiments, the strength grade assessment result is obtained by calculating the ratio of the corrected compressive strength value to the design strength standard value and mapping it to a preset grade range. The preset grade range is divided into a qualified range, a basically qualified range, and an unqualified range, each corresponding to a different grade label. The material construction quality grade assessment is completed by quantitatively analyzing the material uniformity distribution spectrum. Low-uniformity areas in the material uniformity distribution spectrum are segmented by setting a uniformity threshold. The percentage of the total volume of low-uniformity areas relative to the total volume of the concrete component is statistically analyzed, and the average uniformity index value within the low-uniformity areas is calculated. The material construction quality grade is determined by referring to the construction quality grade standard table based on the percentage and the average uniformity index value. The comprehensive structural safety evaluation is generated based on the strength grade assessment result and the material construction quality grade. The comprehensive structural safety evaluation score is calculated using the following formula:

[0103] ;

[0104] in: This indicates the overall score for structural safety evaluation; Indicates the intensity level rating. It is a numerical score converted from the intensity level assessment results; This indicates the material construction quality grade rating. It is a numerical score converted based on the material construction quality grade; and It is a preset weighting coefficient and satisfies Structural safety comprehensive evaluation score This is mapped to a comprehensive evaluation level, which includes safe, basically safe, and unsafe.

[0105] It is understandable that the planning of the repair and reinforcement area directly uses the spatial location and size information of the structural weak area markers. These markers include the geometric center coordinates, equivalent diameter, and volume of each weak area. During planning, the priority of repair is determined based on the spatial distribution density and equivalent diameter of the weak areas. In some embodiments, the maintenance strategy library stores multiple maintenance strategy records. Each record is associated with the comprehensive structural safety evaluation level, characteristic parameters of the repair and reinforcement area, and corresponding construction process descriptions, material specifications, and estimated construction period. The matching process is achieved by inputting the current comprehensive structural safety evaluation level and the characteristic parameters of the repair and reinforcement area into the maintenance strategy library for querying. Optionally, the formation of the maintenance recommendation scheme includes generating a plan layout and section view of the repair and reinforcement area, marking specific construction process steps and material usage lists, and calculating the construction period based on the total volume of the repair and reinforcement area and the standard time quota of the selected construction process. It is understandable that standardized test reports use a fixed template structure. This template includes sections such as project overview, test data, analysis results, comprehensive evaluation, and maintenance recommendations. Strength grade assessment results, material construction quality grades, comprehensive structural safety evaluation, and maintenance recommendations are automatically populated into the corresponding sections, along with relevant data charts. Optionally, the material uniformity distribution spectrum in the test report is presented as a color cloud map, weak structural areas are marked as highlighted areas on the component's 3D model, and maintenance recommendations are listed in a combination of tables and text.

[0106] See Figure 5 This is a bar chart comparing the corrected compressive strength of 10 concrete components. It visually presents the comparison between the strength test results and design standard values ​​of these 10 components, showing significant fluctuations in overall strength. 30% of the components are substandard, while 30% are basically qualified. Special attention needs to be paid to the construction quality and material uniformity of this batch of components. This chart can be directly used in test reports to identify large strength fluctuations and concentrated occurrences of substandard components, indicating the need to trace back the construction process and investigate the causes of quality fluctuations. A graded assessment should clearly distinguish between qualified, basically qualified, and substandard components, providing core data for subsequent structural safety assessments. For basically qualified components, monitoring frequency should be increased to assess long-term load-bearing capacity. Qualified components can be maintained according to the regular schedule. For substandard components, reinforcement or replacement plans must be developed immediately.

[0107] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for testing the compressive strength of concrete in building engineering, characterized in that, Includes the following steps: A multi-dimensional physical field data set of concrete components is collected, which includes ultrasonic propagation data, infrared thermal imaging sequences, and three-dimensional surface topography point clouds. The multi-dimensional physical field data set is subjected to cross-modal fusion processing to generate a feature mapping map of internal defects in concrete and a material homogeneity distribution spectrum. The pre-trained strength prediction model is invoked to analyze the internal defect feature mapping map and material homogeneity distribution spectrum of the concrete, and to generate the predicted compressive strength value and the marker of the weak area of ​​the structure. Based on the marking of the structural weak areas, the predicted compressive strength value is spatially weighted to generate the corrected compressive strength value. Based on the corrected compressive strength value and material uniformity distribution spectrum, a test report and maintenance recommendations are generated. The process of performing cross-modal fusion processing on the multi-dimensional physical field data set to generate a feature mapping map of internal defects in concrete and a material homogeneity distribution spectrum includes: The ultrasonic propagation data is processed by time-frequency analysis to extract the propagation velocity distribution, attenuation coefficient distribution and reflected signal energy distribution of ultrasonic waves inside the concrete, and to generate three-dimensional ultrasonic feature data. The infrared thermal imaging sequence is used to perform thermal flow field inversion calculations to reconstruct the distribution of heat conduction parameters inside the concrete and generate three-dimensional thermodynamic feature data. Structured light feature analysis is performed on the three-dimensional surface topography point cloud to calculate the distribution density and depth features of surface microcracks and generate three-dimensional topography feature data. A unified spatial coordinate reference is established, and the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data are spatially registered and voxel aligned. At each aligned voxel position, the propagation velocity and attenuation coefficient from the three-dimensional ultrasonic feature data, the heat conduction parameter value from the three-dimensional thermodynamic feature data, and the microcrack depth feature from the three-dimensional morphology feature data are fused together. Through feature cascading and normalization, the feature mapping map of the internal defects of the concrete is generated. Based on the consistency between the uniformity of reflected signal energy distribution in the three-dimensional ultrasonic feature data and the thermal conductivity parameters in the three-dimensional thermodynamic feature data, the uniformity distribution spectrum of the material is calculated and generated. The process involves calling a pre-trained strength prediction model to analyze the internal defect feature map and material homogeneity distribution spectrum of the concrete, generating predicted compressive strength values ​​and markers for structurally weak areas, including: The feature map of internal defects in the concrete is input into the defect feature extraction branch of the strength prediction model. The defect feature extraction branch is constructed based on a convolutional neural network and is used to identify and quantify the size, shape and spatial location information of internal pores, cracks and segregation defects. The material uniformity distribution spectrum is input into the uniformity analysis branch of the strength prediction model, and the uniformity analysis branch calculates the coefficient of variation and gradient trend of the material properties in space. The quantitative defect information output from the defect feature extraction branch and the material variation information output from the uniformity analysis branch are combined, and the intensity contribution value of each local region is calculated through a fully connected network layer. The predicted compressive strength is calculated by spatial integration based on the strength contribution values ​​of all local regions. Identify local areas where the intensity contribution value is lower than the global average contribution value by a certain proportion, extract their boundary coordinates, and generate the weak area markers of the structure.

2. The method for testing the compressive strength of concrete in building engineering according to claim 1, characterized in that, The step of performing thermal flow field inversion calculation on the infrared thermal imaging sequence to reconstruct the distribution of heat conduction parameters inside the concrete and generate three-dimensional thermodynamic feature data includes: Temperature field calibration and noise filtering are performed on each frame of the infrared thermal imaging sequence to obtain surface temperature field distribution data in the time series. A three-dimensional heat conduction model of the interior of concrete is established based on the heat conduction equation. The three-dimensional heat conduction model includes initial estimates of thermal conductivity, specific heat capacity and density. The surface temperature field distribution data over the time series is used as boundary conditions to drive the three-dimensional heat conduction model to perform iterative solutions. In each iteration, the thermal conductivity parameter values ​​of the voxels inside the three-dimensional thermal conductivity model are adjusted to minimize the difference between the surface temperature field calculated by the model simulation and the measured surface temperature field distribution data on the time series. When the difference converges to a preset threshold, the iteration stops. At this point, the thermal conductivity parameter value of each voxel in the three-dimensional thermal conductivity model constitutes the three-dimensional thermodynamic feature data.

3. The method for testing the compressive strength of concrete in building engineering according to claim 1, characterized in that, The step of spatially weighting the predicted compressive strength value based on the markers of the structurally weak areas to generate a corrected compressive strength value includes: Based on the marked weak areas of the structure, determine the proportion and spatial distribution of the weak areas in the overall volume of the concrete component; Query the preset weak area influence coefficient table, which defines the correction weights for the predicted compressive strength values ​​of weak areas with different proportions and distribution patterns. Based on the table of influence coefficients for weak areas, obtain the corrected weighting coefficients that match the current proportion and distribution pattern of weak areas; Multiply the predicted compressive strength value by the correction weighting coefficient to obtain the preliminary corrected strength value; Simultaneously, the degree of overlap between the structural weak area markers and the low uniformity areas in the material uniformity distribution spectrum is analyzed, and the preliminary corrected strength value is finely adjusted based on the degree of overlap to generate the corrected compressive strength value.

4. The method for testing the compressive strength of concrete in building engineering according to claim 3, characterized in that, The boundary coordinates of local regions whose identification intensity contribution value is lower than the global average contribution value by a certain proportion are extracted to generate the structurally weak region markers, including: The average value of the intensity contribution values ​​of all local regions output by the intensity prediction model is calculated as the global average contribution value. A proportional threshold is set to identify all local regions whose intensity contribution value is lower than the global average contribution value multiplied by the proportional threshold. Spatial clustering analysis was performed on all identified low-intensity contributing local areas to merge spatially adjacent and similar-attributed areas into the same weak area. For each merged weak region, extract the set of voxel coordinates of its outer contour, and calculate the geometric center coordinates, equivalent diameter, and volume of the weak region; The set of voxel coordinates, geometric center coordinates, equivalent diameter and volume information of each weak region are structurally encapsulated to form the weak region marker.

5. The method for testing the compressive strength of concrete in building engineering according to claim 1, characterized in that, Based on the corrected compressive strength value and material uniformity distribution spectrum, a test report and maintenance recommendation plan are generated, including: The corrected compressive strength value is compared with the design strength standard value to determine the strength grade evaluation result; Analyze the distribution range and severity of low uniformity regions in the material uniformity distribution spectrum to assess the material construction quality grade; Based on the strength grade assessment results and the material construction quality grade, a comprehensive structural safety evaluation is generated. Based on the spatial location and size information of the marked weak areas in the structure, specific repair and reinforcement areas are planned; Based on the comprehensive evaluation of structural safety and the repair and reinforcement areas, the corresponding construction techniques, material requirements and construction period estimates are matched from the maintenance strategy library to form the maintenance recommendation plan; The strength grade assessment results, material construction quality grade, comprehensive evaluation of structural safety, and maintenance recommendations are integrated into a standardized test report.

6. The method for testing the compressive strength of concrete in building engineering according to claim 2, characterized in that, The establishment of a three-dimensional heat conduction model inside concrete based on the heat conduction equation includes: The concrete component is discretized in space into a three-dimensional mesh model composed of regular hexahedral voxels; Each voxel is assigned an initial value for thermal conductivity, specific heat capacity, and density. The initial values ​​are determined based on the concrete mix design and a standard material parameter library. Establish the heat conduction relationship between each voxel and its neighboring voxels, and construct the heat conduction equation set for the entire three-dimensional mesh model based on Fourier's law and the law of conservation of energy. The heat conduction equations take the temperature change of each voxel over time as the unknown quantity and the thermal conductivity, specific heat capacity and density between voxels as parameters. The initial temperature field distribution and external environmental boundary conditions of the model are set, including the air convection heat transfer coefficient and the solar radiation heat flux density.

7. The method for testing the compressive strength of concrete in building engineering according to claim 1, characterized in that, The establishment of a unified spatial coordinate reference, and the spatial registration and voxel alignment of the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data, include: In each dataset, select at least three non-collinear, identifiable common feature points, including component corner points, pre-embedded marker points, or significant surface features; Based on the common feature points, a spatial transformation matrix is ​​calculated to map different data sets to the same coordinate system. The spatial transformation matrix includes a rotation matrix and a translation vector. The spatial transformation matrix is ​​used to perform coordinate transformations on the three-dimensional ultrasonic feature data, the three-dimensional thermodynamic feature data, and the three-dimensional morphological feature data, respectively. Under the transformed unified coordinate system, a common three-dimensional mesh containing the entire concrete component is defined, and the voxel size of the mesh is set to be no greater than the resolution of the minimum data set; The parameter values ​​in each dataset are resampled to the center point of each voxel in the common 3D mesh using cubic spline interpolation to complete voxel alignment.

8. A concrete compressive strength testing system for building engineering, characterized in that, It includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement a method for testing the compressive strength of concrete for building engineering, as described in any one of claims 1 to 7.

Images