Nondestructive testing method for multiple defects in built building structure

Abstract

The invention relates to the technical field of building structure detection, in particular to a nondestructive detection method for multiple defects in a built building structure, which comprises the following steps of: constructing a precision digital twin body through three-dimensional laser scanning, and combining time-space registration of multi-source data such as ultrasonic guided waves and geological radar to obtain a three-dimensional digital twin body; multi-dimensional defect characterization of geometrical morphology-material damage-dynamic response is realized, a time sequence precision index is introduced, a defect development trend is captured in real time through coupling calculation of a steady-state precision weight and a precision change rate, the interference degree is quantitatively detected on the basis of a cumulative damage model (CDM), and through double weighting of a mechanical response interference factor (MRIF) and a detection area, the defect detection accuracy is improved. According to the method, the structure additional damage is effectively analyzed, the periodic optimization of a built building detection scheme is ensured based on the real-time data iteration of the digital twinborn body, and the full-chain technology innovation of digital twinborn modeling-multi-source data fusion-multi-target optimization-dynamic updating is adopted.

Description

Technical Field

[0001] The present invention relates to the technical field of building structure detection, and more particularly to a non-destructive detection method for multiple defects inside an existing building structure. Background Art

[0002] In the field of nondestructive testing for internal defects in existing building structures, traditional technologies typically use a single detection method (such as ultrasonic guided waves, geological radar, or infrared thermal imaging) to identify structural defects. Limited by the single detection principle, these methods struggle to comprehensively cover multiple types of defects, such as cracks, voids, and steel corrosion. Furthermore, they suffer from significant blind spots and insufficient accuracy. For example, ultrasonic guided waves are highly sensitive to deep defects but have low resolution for surface defects. Infrared thermal imaging can only identify near-surface damage, and geological radar is susceptible to interference from metal, resulting in missed detection rates exceeding 30% in multi-defect scenarios. Furthermore, existing multi-technology collaborative testing often remains at the level of independent data analysis, lacking mechanisms for spatiotemporal registration and feature fusion of cross-modal data. This makes it difficult to form a comprehensive defect representation. Furthermore, testing solutions often rely on empirical settings, failing to strike a balance between detection accuracy, efficiency, and structural protection.

[0003] With the increasing aging of building structures and the growing demand for intelligent operations and maintenance, the limitations of traditional inspection methods are becoming increasingly significant: First, they lack the ability to dynamically track the evolution of structural defects, making it impossible to capture defect development trends; second, the interference effect of the inspection process on the structure is not quantified, which may lead to additional damage due to frequent inspections; and third, the formulation of inspection plans lacks a scientific multi-objective optimization framework, which often leads to "over-inspection" or "under-inspection". Therefore, how to build a multi-dimensional inspection system that integrates geometry, material damage, and dynamic response, and realizes intelligent optimization and dynamic updating of inspection plans, has become a key technical bottleneck in solving the problem of accurate multi-defect detection and safe operation and maintenance of existing buildings.

[0004] Based on the above content, the present invention proposes a non-destructive detection method for multiple defects inside the structure of an existing building. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the object of the present invention is to provide a non-destructive detection method for multiple defects inside the structure of an existing building.

[0006] To achieve the above object, the present invention provides the following technical solutions: A nondestructive testing method for multiple defects inside an existing building structure, the method steps are as follows: Step 1: 3D laser scanning, BIM model reconstruction and twin attribute assignment; Step 2: Construction of multi-source detection data system; Step 3: Definition of decision space and objective function; Step 4: Define multi-objective functions, establish a detection scheme-target value mapping database, and accelerate evaluation through the Gaussian process regression model; Step 5: Generate candidate solutions through the NSGA-III algorithm, and finally generate multiple groups of non-dominated solutions to form a Pareto optimal solution set. Select the periodic optimal solution from the Pareto optimal solution set. Step 6: Based on the nondestructive testing cycle, periodically update the cycle optimization plan.

[0007] Furthermore, 3D laser scanning is implemented: a scanner is used to perform multi-site scanning of the building structure, and point cloud stitching is achieved through target balls; blind areas are supplemented with endoscopic laser scanning; BIM model reconstruction: based on the point cloud, the LOD400 precision model is rebuilt in Revit, and the deviation between the component outline and the point cloud is ≤1mm; a hierarchical model architecture is established: geometry layer, component layer, and node layer.

[0008] Furthermore, data spatiotemporal registration and feature extraction, including coordinate unification and cross-modal feature fusion, establish a defect feature association table and form a multi-technology cross-validation rule base.

[0009] Furthermore, the multi-objective function definition includes goal 1: comprehensive detection accuracy index, goal 2: comprehensive detection non-destructive index and goal 3: detection cost.

[0010] Further, the calculation steps of the comprehensive detection accuracy index are: Calculate the comprehensive detection accuracy index , where I is the total number of building defect types and a1 is the time series precision coefficient.

[0011] Further, through Calculate shape matching metrics , where i corresponds to the building defect type, j corresponds to the predicted defect boundary point, is the three-dimensional coordinate vector (xj, yj, zj) of the j-th predicted defect boundary point, J represents the total number of predicted defect boundary points, represents the boundary surface of the i-th building defect type, To predict defect boundary points To Boundary Surface The shortest Euclidean distance.

[0012] Further, through Calculate the time series accuracy index , t corresponds to the detection time, T is the total number of detection times in a nondestructive testing cycle, is the steady-state accuracy weight coefficient, is the defect detection accuracy at the time of detection, is the rate of change of detection accuracy between adjacent detection moments.

[0013] Further, the calculation steps of the lossless index are as follows: Calculate the nondestructive detection index , H represents the total number of inspections of the building structure in one nondestructive testing cycle, K is the number of area divisions in a single inspection, : mechanical response interference factor of the kth region in the hth test, : The area of ​​the kth detection area, A is the total area of ​​the building structure.

[0014] Furthermore, a period-optimal solution is selected from the Pareto optimal solution set: the period-optimal index of each detection solution in the Pareto optimal solution set is obtained, and the detection solution with the largest period-optimal index value is marked as the period-optimal solution.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The method of the present invention constructs a precision digital twin through three-dimensional laser scanning, combines the spatiotemporal registration of multi-source data such as ultrasonic guided waves and geological radar, realizes the multi-dimensional defect characterization of "geometry-material damage-dynamic response", introduces time series accuracy indicators, and captures the defect development trend in real time through the coupled calculation of steady-state accuracy weight and accuracy change rate. The detection interference degree is quantified based on the cumulative damage model (CDM), and the additional structural damage is effectively analyzed through the dual weighting of the mechanical response interference factor (MRIF) and the detection area. Based on the real-time data iteration of the digital twin, the periodic optimization of the detection plan for existing buildings is guaranteed. Through the full-chain technical innovation of "digital twin modeling-multi-source data fusion-multi-objective optimization-dynamic update", the contradictions in traditional non-destructive testing in terms of accuracy, efficiency and structural protection are broken through, and a technological leap from "single-point detection" to "intelligent diagnosis" is achieved, providing a complete solution for the safe operation and maintenance of existing buildings that is both scientific and engineering practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention is a flowchart of a method for non-destructive testing of multiple defects inside an existing building structure; Figure 2 Build a flow chart for multi-source detection data system; Figure 3 Define a flow graph for a multi-objective function. DETAILED DESCRIPTION

[0017] Reference Figures 1 to 3 , a non-destructive testing method for multiple defects inside the structure of an existing building, the method steps are as follows: Step 1: 3D laser scanning and BIM model reconstruction. 3D laser scanning implementation: A Faro Focus S350 scanner (point cloud density 100 points / cm²) was used to perform multi-site scanning of the building structure. Point cloud stitching was achieved using a target sphere (positioning accuracy ±0.1mm), and the global control network error was ≤0.5mm. Complex nodes (such as domes and special-shaped columns) were scanned using an Artec Eva handheld scanner (resolution 0.1mm), and blind areas were supplemented with endoscopic laser scanning. BIM model reconstruction: Based on the point cloud, an LOD400 precision model was reconstructed in Revit, with the deviation between the component outline and the point cloud ≤1mm. A hierarchical model architecture was established: geometry layer (precise outline), component layer (beam / column / slab classification), and node layer (connection details).

[0018] Twin properties are assigned and material parameters are mapped: concrete strength grade is measured using the rebound method (with an error of ±5%), with elastic modulus E = 30 GPa @ C30 and density ρ = 2400 kg / m³; steel properties are based on factory reports, with yield strength f_y = 345 MPa and Poisson's ratio ν = 0.3. Defect pre-implantation simulation: Typical defects (such as a 5 cm × 10 cm cavity and a 0.2 mm crack) are virtually implanted in the twins. Finite element simulation is performed using ANSYS to generate a library of ultrasonic and radar signal signatures corresponding to the defects.

[0019] Step 2: Construct a multi-source detection data system (aims to convert heterogeneous data collected by different technologies (modalities) such as ultrasonic guided waves, infrared thermal imaging, and geological radar into a unified, standardized, and fusible standard data format). Standardized detection data collection: The following examples are four technology types, which are not exhaustive: Technology type: ultrasonic guided waves; Equipment model: OmniScan MX2; Acquisition parameters: 5MHz phased array, sampling rate 10MS / s, 64 array elements; Data format: .dat (raw waveform); Technology type: geological radar; Equipment model: SIR-4000; Acquisition parameters: 400MHz antenna, 2048 sampling points / lines, time window 50ns; Data format: .rd3 (3D data volume); Technology type: infrared thermal imaging; Equipment model: FLIR T1040; Acquisition parameters: 640 × 512 pixels, thermal sensitivity 20 mK; Data format: .tiff (temperature matrix); Technology type: Laser vibrometer; Device model: OFV-505; Acquisition parameters: Sampling rate 100 kHz, scanning interval 5 cm; Data format: .csv (vibration spectrum).

[0020] Data spatiotemporal registration and feature extraction, coordinate unification (binding all extracted features (such as ultrasonic echo attenuation values) to three-dimensional spatial coordinates to form a "feature value-coordinate" mapping table), cross-modal feature fusion: establishing a defect feature association table, such as associating "ultrasonic echo attenuation 20dB + radar reflection intensity -15dB" with "diameter > 3cm cavity", forming a multi-technology cross-validation rule base.

[0021] Step 3: Definition of decision space and objective function, three-dimensional encoding of decision variables, technical sequence variables: permutation coding is used, such as [2,1,3,4] to represent the detection order of "infrared → ultrasonic → radar → laser", and the search space is 4!=24 combinations; parameter variables: ultrasonic scanning spacing d∈[5,20]cm (continuous variable), sampling frequency f∈{5,7.5,10}MHz (discrete variable); radar antenna frequency f∈{100,400,900}MHz, scanning line spacing s∈[5,15]cm; spatial strategy variables: generate the detection area priority matrix, such as beam-column node weight w∈[0.8,1.0], wall w∈[0.5,0.7].

[0022] Step 4: Multi-objective function definition, objective 1: comprehensive detection accuracy index, calculation steps of comprehensive detection accuracy index: through Calculate the comprehensive detection accuracy index , where I is the total number of building defect types, a1 is the time series precision coefficient, and the value of a1 can be 0.4.

[0023] pass Calculate shape matching metrics , where i corresponds to the building defect type (e.g., when inspecting a concrete column, the building defect type may be a main crack, a honeycomb void, or a steel bar corrosion area), and j corresponds to the predicted defect boundary point. is the three-dimensional coordinate vector (xj, yj, zj) of the j-th predicted defect boundary point, which is generated by the defect recognition results of detection technology (such as ultrasound, radar), and J represents the total number of predicted defect boundary points. The boundary surface of the i-th building defect type is a closed manifold in three-dimensional space (such as a crack surface or a cavity wall). To predict defect boundary points To Boundary Surface The shortest Euclidean distance, which measures the local matching accuracy, is in millimeters (mm).

[0024] pass Calculate the time series accuracy index , t corresponds to the detection time, T is the total number of detection times in a nondestructive testing cycle, is the steady-state accuracy weight coefficient, The value can be 0.6, is the defect detection accuracy at the time of detection, is the rate of change of detection accuracy between adjacent detection moments.

[0025] Objective 2: Comprehensively detect non-destructive indicators. Calculation steps for detecting non-destructive indicators: Calculate the nondestructive detection index H represents the total number of inspections of the building structure during a nondestructive testing cycle. For example, if a bridge has been inspected three times, then H = 3. This quantifies the cumulative effect of inspections during a nondestructive testing cycle. The larger H is, the higher the risk of fatigue damage. K is the number of area divisions in a single inspection (dimensionless). Physical meaning: K means that each inspection divides the building structure into K independent areas, each corresponding to different inspection parameters and interference intensity. Division principle: Division by component type (such as beams, columns, and plates); Example: A building inspection is divided into 10 areas (K = 10), including 2 areas for beam-column joints and 8 areas for walls. : The mechanical response interference factor of the kth region in the hth test (dimensionless), core definition: ,in, : local stress caused by the h-th detection in region k, : local strain caused by the h-th detection in region k; 、 : Yield stress and strain threshold of materials in the k region; : The area of ​​the kth detection area (unit: m²); Function: Characterizes the interference space range of the kth area in a single detection, and reflects the comprehensive impact of "interference intensity × range" after coupling with MRIF. Standardize the area (A is the total area of ​​the structure) to eliminate the influence of structural scale differences. For example: a small area with high risk ( =1m², A=100m², =0.01); large areas of low risk areas ( =50m², =0.5). A is the total area of ​​the building structure: used as a benchmark for area normalization to make CDM values ​​comparable across structures, A = the total area calculated based on the structural dimensions, or the total area of ​​key components of interest for testing (e.g., only the area of ​​load-bearing components is calculated).

[0026] Objective 3: To measure the cost of testing, we established a database of testing scheme-target value mapping and accelerated the evaluation through the Gaussian process regression (GPR) model.

[0027] Step 5: Generate candidate solutions using the NSGA-III algorithm, ultimately generating multiple groups of non-dominated solutions to form a Pareto optimal solution set. Obtain the cycle optimization index for each inspection solution in the Pareto optimal solution set. (Each inspection solution has different decision variables, such as the technology sequence: [ultrasound → radar → infrared]: suitable for deep defect detection, which is time-consuming; [infrared → laser → ultrasonic]: suitable for initial screening of surface defects, which is more efficient; parameter setting differences: ultrasonic frequency 10MHz (high resolution, suitable for microcracks) vs. 5MHz (deep penetration, suitable for large cavities); spatial strategy differences: beam-column node weight 1.0 (focused scanning, 20% more inspection points) vs. weight 0.8 (conventional scanning). For example, if the Pareto optimal solution set contains inspection plan A and inspection plan B, the technology sequence variables for inspection plan A are: [1, 3, 2, 4]. →Ultrasonic guided waves → Geological radar → Infrared thermal imaging → Laser vibrometer; Parameter variables: Ultrasonic scanning spacing d = 5 cm (minimum spacing), sampling frequency f = 10 MHz (maximum frequency); Radar antenna frequency f = 900 MHz (maximum frequency), scanning line spacing s = 5 cm (minimum spacing); Spatial strategy variables: Beam-column node weight w = 1.0 (highest priority), wall weight w = 0.6 (medium priority); Technical sequence logic: First, use ultrasonic guided waves (10 MHz) to detect internal cracks in concrete (resolution ≤ 0.2 mm), then use 900 MHz radar to precisely locate voids (depth ≤ 50 cm, resolution 3 cm), and finally use Infrared and laser vibrometers verify surface defects and dynamic response, forming a progressive inspection chain from "interior-near-surface-dynamic." Parameter selection impacts: Ultrasonic 5cm spacing + 10MHz frequency: While sacrificing detection efficiency (scanning speed reduced by 40%), it enables precise location of microcracks (0.1mm level); Radar 900MHz high frequency: Penetration depth is reduced to 1.5m (50% reduction compared to 100MHz), but cavity boundary recognition accuracy is improved to 2cm. Spatial strategy highlights: A weight of 1.0 for beam-column nodes means the inspection point density in this area is doubled compared to walls (e.g., scanning points per square meter increase from 200 to 400), prioritizing the inspection accuracy of load-bearing components.Detection Plan B's technical sequence variables are: [2, 4, 3, 1] → infrared thermal imaging → laser vibrometer → geological radar → ultrasonic guided waves; parameter variables: ultrasonic scanning spacing d = 20 cm (maximum spacing), sampling frequency f = 5 MHz (minimum frequency); radar antenna frequency f = 100 MHz (minimum frequency), scanning line spacing s = 15 cm (maximum spacing); spatial strategy variables: beam-column node weight w = 0.8 (medium priority), wall weight w = 0.5 (lowest priority); technical sequence logic: first, use infrared thermal imaging (640 × 512 pixels) to quickly scan large-area surface defects (efficiency 1000 m2 / h), then use laser vibrometer to evaluate the structural dynamic characteristics (identify support damage), and finally use low-frequency radar (100 MHz) and ultrasound (5 MHz) to roughly screen suspicious areas, forming an efficient "initial screening-localization-verification" process. Parameter selection impacts: Ultrasonic scanning with a 20cm spacing and a 5MHz frequency increases scanning speed by fourfold (from 10m2 / h to 40m2 / h), but crack detection accuracy drops to 1mm. Radar with a 100MHz low-frequency achieves a penetration depth of 3m (a 100% increase compared to 900MHz), but cavity resolution drops to 10cm. Spatial strategy focuses on reducing the weight difference between beam-column joints and walls (0.8 vs. 0.5), reducing the density of inspection points, and making it suitable for uniform inspection of large structures. The inspection solution with the highest period optimization index value is marked as the period optimization solution.

[0028] Step 6: Based on the nondestructive testing cycle (the duration of the nondestructive testing cycle is comprehensively set based on the structural characteristics and engineering constraints of the existing building), periodically update the cycle optimization plan (because the modeling data in the digital twin of the existing building is updated, the cycle optimization plan needs to be updated regularly).

[0029] Steps for obtaining the cycle optimization index of the detection scheme: Obtain the comprehensive detection accuracy index of a detection scheme , comprehensive testing non-destructive indicators Testing costs ,pass Calculate the cycle optimization index of the detection scheme (Dimensionless calculation), where a2 is the comprehensive detection accuracy coefficient, a3 is the comprehensive detection non-destructive coefficient, and a4 is the detection cost coefficient. The value of a2 can be 0.4, the value of a3 can be 0.3, and the value of a4 can be 0.3.

[0030] The above method constructs a precise digital twin through 3D laser scanning, and combines the spatiotemporal registration of multi-source data such as ultrasonic guided waves and geological radar to achieve multi-dimensional defect characterization of "geometry-material damage-dynamic response". It introduces time series accuracy indicators and captures defect development trends in real time through the coupled calculation of steady-state accuracy weight and accuracy change rate. It quantifies the detection interference degree based on the cumulative damage model (CDM). The dual weighting of the mechanical response interference factor (MRIF) and the detection area is used to effectively analyze the additional structural damage. Based on the real-time data iteration of the digital twin, the periodic optimization of the inspection plan for existing buildings is guaranteed. Through the full-chain technological innovation of "digital twin modeling-multi-source data fusion-multi-objective optimization-dynamic update", it breaks through the contradictions of traditional non-destructive testing in terms of accuracy, efficiency and structural protection, and realizes the technological leap from "single-point detection" to "intelligent diagnosis", providing a complete solution that combines scientificity and engineering practicality for the safe operation and maintenance of existing buildings.

[0031] The above formulas are all dimensionless and numerically calculated, and the preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0032] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0033] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0034] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0035] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0036] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0037] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0038] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A non-destructive testing method for multiple defects inside an existing building structure, characterized in that: The steps are as follows: Step 1: 3D laser scanning, BIM model reconstruction and twin attribute assignment; Step 2: Construction of multi-source detection data system; Step 3: Definition of decision space and objective function; Step 4: Define multi-objective functions, establish a detection scheme-target value mapping database, and accelerate evaluation through the Gaussian process regression model; Step 5: Generate candidate solutions through the NSGA-III algorithm, and finally generate multiple groups of non-dominated solutions to form a Pareto optimal solution set. Select the periodic optimal solution from the Pareto optimal solution set. Step 6: Based on the nondestructive testing cycle, periodically update the cycle optimization plan.

2. The nondestructive testing method for multiple defects inside an existing building structure according to claim 1, characterized in that: Implementation of 3D laser scanning: Using a scanner, perform multi-site scanning of the building structure and achieve point cloud stitching through a target sphere; supplement blind area endoscopic laser scanning; BIM model reconstruction: Reconstruct the LOD400 precision model in Revit based on the point cloud, with the deviation between the component outline and the point cloud ≤1mm; establish a hierarchical model architecture: geometry layer, component layer, and node layer.

3. The nondestructive testing method for multiple defects inside an existing building structure according to claim 1, characterized in that: Data spatiotemporal registration and feature extraction, including coordinate unification and cross-modal feature fusion, establish defect feature association table, and form a multi-technology cross-validation rule base.

4. The nondestructive testing method for multiple defects inside an existing building structure according to claim 1, characterized in that: The multi-objective function definition includes goal 1: comprehensive detection accuracy index, goal 2: comprehensive detection non-destructive index and goal 3: detection cost.

5. The nondestructive testing method for multiple defects inside an existing building structure according to claim 4, characterized in that: Calculation steps for comprehensive detection accuracy index: Calculate the comprehensive detection accuracy index , where I is the total number of building defect types and a1 is the time series precision coefficient.

6. The nondestructive testing method for multiple defects inside an existing building structure according to claim 5, characterized in that: pass Calculate shape matching metrics , where i corresponds to the building defect type, j corresponds to the predicted defect boundary point, is the three-dimensional coordinate vector (xj, yj, zj) of the j-th predicted defect boundary point, J represents the total number of predicted defect boundary points, represents the boundary surface of the i-th building defect type, To predict defect boundary points To Boundary Surface The shortest Euclidean distance.

7. The nondestructive testing method for multiple defects inside an existing building structure according to claim 5, characterized in that: pass Calculate the time series accuracy index , t corresponds to the detection time, T is the total number of detection times in a nondestructive testing cycle, is the steady-state accuracy weight coefficient, is the defect detection accuracy at the time of detection, is the rate of change of detection accuracy between adjacent detection moments.

8. The nondestructive testing method for multiple defects inside an existing building structure according to claim 4, characterized in that: Calculation steps for detecting non-destructive indicators: Calculate the nondestructive detection index , H represents the total number of inspections of the building structure in one nondestructive testing cycle, K is the number of area divisions in a single inspection, : mechanical response interference factor of the kth region in the hth test, : The area of ​​the kth detection area, A is the total area of ​​the building structure.

9. The nondestructive testing method for multiple defects inside an existing building structure according to claim 1, characterized in that: Select the period-optimal solution from the Pareto optimal solution set: obtain the period-optimal index of each detection solution in the Pareto optimal solution set, and mark the detection solution with the largest period-optimal index value as the period-optimal solution.

Images