Method for detecting and analyzing carbon fiber cloth reinforced damaged concrete structure

Abstract

The invention discloses a carbon fiber cloth reinforced damaged concrete structure detection and analysis method, and relates to the technical field of concrete structure health monitoring, and the method comprises the steps: extracting a high-risk point set from a three-dimensional damage distribution map, carrying out the weighted K-means clustering to construct a Voronoi subspace, calculating the failure probability, and dividing an emergency repair block set; generating a processing queue by using the repairing blocks according to priorities, planning a spiral path of the plasma spray gun by adopting a TSP algorithm, and generating a surface roughness distribution diagram in combination with laser confocal and white light interference detection; according to the roughness partition matching epoxy resin-carbon fiber dynamic matching scheme, a pressure thermodynamic diagram is generated through finite element simulation and electrode time-sharing electrification. Through fusion of the damage evolution graph and the digital twin model, the plasma etching parameters and the conductivity of the carbon fiber cloth are dynamically adjusted, so that the tensile strength of the repaired area is improved, and full-life-cycle health monitoring is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring of concrete structures, and particularly to a method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth. Background Art

[0002] In recent years, the health monitoring technology of concrete structures mainly relies on means such as acoustic emission detection, vibration frequency analysis, and infrared thermal imaging. The traditional acoustic emission technology evaluates damage by collecting stress wave signals, but due to the locality of single-point excitation, it is difficult to achieve large-scale high-resolution detection; although vibration frequency analysis can reflect the overall stiffness degradation of the structure, it is insufficient in sensitivity to microcracks. In the prior art, a laser-ultrasonic composite detection method is mostly adopted, in which acoustic emission signals are excited by pulsed lasers and dynamic responses are obtained by combining vibration sensors, but there are still technical bottlenecks in the characterization of damage evolution under dynamic loads and the fusion of multi-source data.

[0003] The deficiencies of the current technology are mainly reflected in two aspects: First, most existing detection systems adopt a single laser source or a sequential triggering mechanism, resulting in insufficient spatio-temporal synchronization of acoustic emission signals and vibration responses, which affects the correlation analysis between transient stress waves and material dynamic characteristics; Second, traditional damage assessment models rely on static parameters (such as amplitude attenuation values), lack the quantitative fusion of dynamic propagation rates and three-dimensional damage fields, and it is difficult to construct an accurate mapping between damage evolution and structural failure risks. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth to solve the problems of insufficient spatio-temporal synchronization of multi-source data and difficulty in quantitative fusion of dynamic propagation rates and three-dimensional damage fields in concrete structure damage detection.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth, which includes performing grid scanning on the concrete surface using pulsed laser and continuous laser, synchronously collecting amplitude attenuation value and frequency offset data, performing fusion processing, and generating a three-dimensional damage distribution map; Extracting a high-risk point set from the three-dimensional damage distribution map, performing weighted K-means clustering to construct a Voronoi subspace, calculating the failure probability through Monte Carlo simulation, and dividing an emergency repair block set; Generating a processing queue for the repair blocks according to the priority, planning a spiral path of a plasma spray gun using the TSP algorithm, and combining laser confocal and white light interference detection to generate a surface roughness distribution map; According to the roughness zoning, the dynamic ratio of epoxy resin to carbon fiber is matched, and the pressure thermodynamic diagram is generated through finite element simulation and electrode time-sharing power supply; Deploy fiber optic sensing networks to decouple strain-temperature data, build and train damage development prediction models, predict remaining life, and generate structural monitoring assessment reports.

[0007] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, wherein: the generation of a three-dimensional damage distribution map comprises the following steps: Calibrate the energy parameters of pulsed lasers and continuous lasers, and plan the scanning path; Synchronously stimulate the acoustic emission signal and measure the frequency offset, align the amplitude attenuation value and the frequency offset data stream, and generate the amplitude matrix and frequency matrix after removing the abnormal values; Based on the initial amplitude sequence of the impact center point, a discrete depth matrix is ​​generated by integrating the logarithmic decay rate. The three-dimensional damage distribution map is output through bicubic spline interpolation and the high-risk areas are marked.

[0008] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, the calculation of failure probability and division of emergency repair block set includes the following steps: A set of high-risk points is extracted from the three-dimensional damage distribution map, weighted based on the damage depth and extension rate, and the Voronoi subspace is generated using weighted K-means clustering. Inject normal distribution random perturbations into the subspace and calculate the cumulative failure probability; The subspace weights and cumulative failure probabilities are integrated to generate a failure weight matrix, and DBSCAN clustering is performed to generate a set of emergency repair blocks sorted by priority.

[0009] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, wherein: the generation of the surface roughness distribution map includes the following steps: The radius of the repair block set is expanded according to the priority and the overlapping area is detected, and the spiral path is generated using the TSP algorithm; Match the plasma power and helium-oxygen ratio according to the damage depth of the repaired area, write the G code and adjust the etching parameters in real time; Laser confocal microscope is used to detect the etching depth and make additional etching, and white light interferometer scanning is used to generate a surface roughness distribution map.

[0010] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, wherein: the generation of the pressure thermodynamic map includes the following steps: The area is divided according to the roughness distribution map, and the viscosity of epoxy resin and the proportion of carbon nanotubes are matched to generate a dynamic ratio scheme; After spraying the primer, the coating pressure is adjusted based on the penetration depth feedback, and the expected strain distribution is generated by simulation; The carbon fiber cloth is cut and the electrodes are arranged. The temperature-strain curve is generated by power-on at different times, and the pressure thermodynamic diagram is output by superimposing finite element data.

[0011] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, wherein: the construction of the damage development prediction model includes the following steps: The coordinates of high stress areas are extracted from the pressure thermodynamic map, and the optical fiber sensing network is deployed along the principal stress direction; Decouple the strain and temperature data of the optical fiber sensor, eliminate the interference of thermal expansion through dynamic compensation, and generate the temperature-compensated strain matrix; The strain matrix and the three-dimensional damage map are aligned in time and space to construct a damage development prediction model.

[0012] As a preferred solution of the carbon fiber cloth reinforced damaged concrete structure detection and analysis method of the present invention, the training of the damage development prediction model includes the following steps: The damage development prediction model is segmented into time windows, and the strain change increment is calculated voxel by voxel and converted into damage increment; The spatiotemporal convolutional neural network is used to extract the damage evolution characteristics, and the damage prediction model is trained with finite element simulation data. The generative adversarial network is used to simulate multi-scenario aging paths, and the mapping relationship between carbon fiber cloth performance degradation and damage extension is established.

[0013] As a preferred solution of the method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to the present invention, the generation of a structural monitoring and evaluation report refers to building a full life cycle health monitoring platform for the reinforced structure through digital twin technology based on a trained damage development prediction model, outputting the predicted value of the remaining life of the structure and the priority maintenance decision in real time, and generating a structural monitoring and evaluation report.

[0014] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth as described in the first aspect of the present invention is implemented.

[0015] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth as described in the first aspect of the present invention is implemented.

[0016] The beneficial effects of the present invention are as follows: By overlapping the foci of pulsed laser and continuous laser and the optoelectronic synchronous triggering mechanism, it ensures the alignment of the time stamps of acoustic emission signals and vibration frequency offsets, improves the correlation accuracy between the propagation characteristics of transient stress waves and the dynamic response of materials, and solves the problem of phase distortion caused by signal delay in traditional methods; uses bicubic spline interpolation to convert the discrete depth matrix into a three-dimensional damage distribution map, combines hydraulic impact to simulate crack propagation behavior under dynamic loads, and optimizes the subspace probability distribution through GF deviation, significantly improving the spatial resolution of damage depth and propagation rate; through the fusion of the damage evolution map and the digital twin model, dynamically adjusts the plasma etching parameters and the electrical conductivity of carbon fiber cloth, improves the tensile strength of the repair area, and increases the strain decoupling accuracy of the fiber optic sensing network, realizing full-life cycle health monitoring. Description of the Drawings

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is the flowchart of data acquisition and feature extraction in Embodiment 1.

[0019] Figure 2 It is the schematic diagram of risk assessment and probability modeling in Embodiment 1.

[0020] Figure 3 It is the schematic diagram of repair path planning and implementation in Embodiment 1.

[0021] Figure 4 It is the schematic diagram of health monitoring and life prediction in Embodiment 1. Detailed Embodiments

[0022] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention in conjunction with the drawings in the specification.

[0023] Many specific details are set forth in the following description in order to fully understand the present invention, but the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0025] Example 1, reference Figures 1 - 4 , which is the first embodiment of the present invention, provides a method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth, comprising the following steps: S1. Start the pulse laser and the continuous laser, calibrate the energy density (pulse width 10ns) and power respectively, adjust the laser incident angle through the optical platform to ensure that the two laser foci coincide with the same detection point on the concrete surface, trigger the signal acquisition module synchronously, and set the grid scanning interval to 5mm; based on the calibrated laser parameters, establish a rectangular coordinate system on the concrete surface (the origin is the lower left corner of the detection area), plan the horizontal (X-axis) and vertical (Y-axis) scanning paths through motion control, with a horizontal step of 5mm and a vertical step of 5mm, generate a coordinate matrix containing all scanning points, and convert it into a CNC file; load the CNC file to the laser scanning platform, start the pulse laser including energy density and pulse width and the continuous laser including power, trigger the dual lasers through the photoelectric synchronization signal, ensure that the two laser foci coincide with the same scanning point, and obtain the synchronous trigger signal and the laser positioning data of the first point coordinates (X=0, Y=0).

[0026] Move the laser head to each grid point (such as X=5mm, Y=5mm) in the order of the CNC file, and stay at each grid point for 100ms. Use pulsed laser to stimulate acoustic emission signals to bombard the concrete surface to generate transient stress waves and collect amplitude attenuation values. Use the Doppler effect in the vibration sensor of the continuous laser to measure the surface vibration frequency offset. Align the amplitude attenuation value with the frequency offset recorded by the vibration sensor according to the timestamp to generate an instantaneous raw data stream arranged in time series. Perform spatial mapping on the raw data stream, associate the scanning point coordinates with the time series, and generate a three-dimensional index. Fill in the amplitude attenuation value and frequency offset respectively to form an amplitude matrix and a frequency matrix. Use CRC check to eliminate abnormal values ​​in the amplitude matrix and frequency matrix, and output a structured data file.

[0027] Based on the scanning path data in the structured data file, set the impact point coordinates of the hydraulic impact device to the geometric center of the scanning area, configure the impact energy and impact rate (controlled by adjusting the opening of the hydraulic valve), and verify whether the rate parameter is accurate through real-time feedback from the pressure sensor. The output is a calibrated control instruction set for the impact device; execute the impact action according to the control instruction set of the impact device. At the same time, start the acoustic emission sensor and vibration sensor, collect the initial amplitude of the acoustic emission signal at the moment of impact, and continuously record the amplitude data of the time series according to the sampling rate to generate the original waveform data stream with time stamp marks; align the impact data stream with the amplitude matrix and frequency matrix according to the time stamp, insert the impact mark segment, and remove the high-frequency noise through the Butterworth low-pass filter to obtain the filtered data matrix with impact marks.

[0028] Extract the initial amplitude of the time series corresponding to the impact center point coordinates from the filtered data matrix with impact marks, and intercept the instantaneous amplitude sequence; perform logarithmic operations on the initial amplitude and instantaneous amplitude sequence at each time point to generate a continuous logarithmic decay rate curve, and fill the small discontinuities caused by the sampling interval through cubic spline interpolation to obtain a smooth logarithmic decay rate curve data file; numerically integrate the smooth logarithmic decay rate curve data using the trapezoidal rule, divide the integration interval into multiple micro-segments, and calculate the integration contribution of each micro-segment; accumulate the integration contributions of each grid point segment by segment to generate a discrete depth matrix; use the bicubic spline interpolation algorithm to convert the discrete depth matrix into a three-dimensional depth loss depth field to generate a three-dimensional damage distribution map; set the damage depth threshold according to the crack control requirements, and set the expansion rate threshold based on the correlation between the crack propagation rate and structural failure under impact load. When the damage depth is greater than the damage depth threshold and the expansion rate is greater than the expansion rate threshold, automatically mark it as a high-risk area in the three-dimensional damage distribution map.

[0029] S2. Extract all high-risk coordinate point sets from the three-dimensional damage distribution atlas, use the K-means clustering algorithm to perform density adaptive partitioning on the high-risk areas, and assign weights to each high-risk coordinate point based on the damage depth and expansion rate; divide the point set into 512 clusters through weighted K-means, and use the center of each cluster as the seed point of the Voronoi subspace. For the edge subspace, use the mirror extension method to expand virtual points for boundary correction to obtain the seed point coordinate matrix; based on each seed point, construct a three-dimensional Voronoi grid through the Fortune algorithm, where each subspace is defined as the set of all points closest to the seed point; assign the discrete damage points in the discrete depth matrix to the nearest subspace, and count the damage parameters including the number of damage points, average damage depth, average expansion rate, and subspace damage fluctuation coefficient in each subspace; calculate the initial probability distribution according to the number of damage points in each subspace; combine the initial probability distribution and the subspace seed point coordinates to determine the geometric center of the damage area; define the GF deviation based on the geometric center of the damage area, subspace seed points, and initial probability distribution; use the Metropolis-Hastings algorithm to iteratively adjust the position of the seed points to minimize the GF deviation and obtain the optimized subspace probability distribution.

[0030] Furthermore, the geometric center of the damage area is obtained by considering all high-risk damage points as a whole and calculating the average position of all points. The geometric center of the damage area represents the area with the most intensive damage and the highest risk; the distribution of subspace seed points is obtained by dividing high-risk points into multiple subspaces through the weighted K-means algorithm, and each subspace is represented by a seed point (subspace center), where the position of the seed point reflects the concentration trend of damage points in that subspace; the initial probability distribution is that each subspace is assigned an initial probability value according to the number or density of damage points it contains. The higher the probability value, the more severe or critical the damage in that subspace, and it should be given priority during repair.

[0031] The GF deviation is defined by calculating the spatial distance between the geometric center of the damage area and the subspace seed points. The farther the distance, the more severe the deviation of the subspace from the damage core area. Combining with the initial probability value of the subspace, the distance deviation is weighted. The higher the probability of the subspace, the greater the impact of the distance deviation on the GF deviation.

[0032] Normalize the optimized subspace probability distribution to the interval [0, 1] as the initial sampling probability for each subspace, and dynamically adjust the Monte Carlo sample size for each subspace; generate random perturbation values that follow a normal distribution according to the subspace fluctuation coefficient to ensure that the perturbation amplitude matches the subspace characteristics; generate normal distribution perturbation values based on 10% of the average subspace expansion rate to limit the perturbation amplitude to avoid non-physical deviations; generate random samples for each subspace, where each group of samples contains a combination of average damage depth + random perturbation and average expansion rate + random perturbation; when the damage depth of the sample exceeds the damage depth threshold or the expansion rate exceeds the expansion rate threshold, it is marked as a failed sample; count the number of failed samples in each subspace and calculate the failure probability, with the expression: ; where, represents the cumulative failure probability of the -th subspace, quantifying the overall risk of failure of this subspace under dynamic loads, represents the subspace index, represents the index of the Monte Carlo sample, represents the total number of Monte Carlo samples, represents the cumulative distribution function of the standard normal distribution, represents the -th Monte Carlo sample damage value (such as crack depth) in the subspace, represents the failure threshold, and if it exceeds the failure threshold, it is determined to be a failure, represents the damage standard deviation of the subspace, reflecting the damage volatility.

[0033] Fuse the cumulative failure probability with the optimized subspace probability distribution to generate a subspace failure weight matrix; sort the subspace failure weight matrix in descending order, and based on the high-risk area, set a failure weight threshold, extract the subspaces with subspace failure weights greater than the failure weight threshold, and mark them as emergency repair areas; perform DBSCAN clustering on adjacent high-weight subspaces to generate repair blocks; for each repair block, perform weighted fusion to generate a total weight, and rank the blocks according to the total weight to obtain a set of repair blocks with priorities.

[0034] S3. Arrange the set of repair blocks with priorities in descending order of priority levels (level 1 is the highest) for the repair blocks to generate a processing queue, ensuring that high-priority blocks are processed first; expand the processing radius proportionally according to the total block weight. For example, a block with a radius of 5 mm and a total weight of 0.9 is expanded to 5.9 mm, and detect whether the expanded block radius overlaps with other blocks. If there is an overlap, automatically reduce the radius to the critical value (such as 80% of the adjacent block spacing). For high-priority blocks where overlap cannot be avoided, mark them as "forced coverage"; use the Traveling Salesman Algorithm (TSP) to plan the movement path of the spray gun, continuously process adjacent blocks in sequence to reduce idle travel, and obtain an optimized processing queue file (including coordinates, radius, and processing order). Adjust the plasma power according to the average damage depth of the blocks, with the greater the damage depth, the higher the power; for blocks with a damage depth ≥ 1.5 mm, use a helium-oxygen ratio of 2:1 to enhance oxidation, and maintain a ratio of 3:1 for the remaining blocks; with the block center as the center of the circle, generate an equidistant spiral path (step size 0.5 mm) according to the expanded radius and write it into the G-code file; calculate the block area based on the expanded radius in the spiral path and insert a speed instruction into the G-code; map the power, nitrogen-oxygen ratio, and speed instruction of each block to the configuration file; start the plasma spray gun control mechanism, check the helium supply pressure (≥ 0.5 MPa), the status of the oxygen flow sensor, and the accuracy of the infrared temperature measurement module, read the power value, nitrogen-oxygen ratio, and speed instruction in the configuration file, and at the same time load the G-code file to verify the matching of the spiral path coordinates with the block center to ensure no out-of-bounds or duplicate paths.

[0035] The robotic arm positions the spray gun at the center point of the first block 10 cm away from the concrete surface, adjusts the gas flow according to the configured helium-oxygen ratio (such as 2:1), and moves the spray gun uniformly along the spiral path while synchronously activating the plasma to start surface etching; the infrared temperature measurement module collects the surface temperature and transmits it to the PID controller in real time; if the temperature ≤ 38°C, maintain the current power. If 38°C < temperature ≤ 40°C, reduce the power by 5% (such as 1.8 kW → 1.71 kW). If 40°C < temperature ≤ 45°C, reduce the power by 10% (such as 1.8 kW → 1.62 kW), and at the same time increase the helium flow by 5% (25 → 26.25 L / min). If the temperature > 45°C: Immediately stop etching and trigger an alarm; for every progress signal of etching 3 blocks (such as blocks 1 - 3 completed), the spray gun returns to the safe position, starts the laser confocal microscope to scan the surface of the current block, scans 5 sampling points (center + four quadrants), calculates the average etching depth. If the average depth is out of tolerance (for example, greater than 32 μm or less than 28 μm), mark this block as needing re-etching, and mark it as passed if it is qualified; generate a local spiral path for re-etching the out-of-tolerance blocks and generate a block depth detection report and a re-etching path file; after etching all blocks, the white light interferometer scans along the processing path to generate a surface roughness distribution map.

[0036] S4. Divide the concrete surface into high-roughness area, medium-roughness area and low-roughness area according to the surface roughness distribution map, and mark them with different colors (red - high, yellow - medium, green - low) to generate a zoning heat map; for the high-roughness area, use high-permeability epoxy resin with viscosity ≤ 200 cP, 5% carbon nanotubes (to enhance conductivity) and 1.5% microcapsules (to repair microcracks); for the medium-roughness area, use standard epoxy resin with viscosity 300 - 400 cP, 3% carbon nanotubes and 1% microcapsules; for the low-roughness area, match the material properties through low-viscosity epoxy resin with viscosity ≥ 500 cP and 1% carbon nanotubes (only for basic reinforcement); associate the coordinates of each area with the material property matching results to generate a dynamic ratio mapping table.

[0037] Further explanation, the high-roughness area is the red heat value, corresponding to the deep damage area; the medium-roughness area is the yellow transition area, corresponding to moderate damage; the low-roughness area is the green background area, corresponding to the slightly damaged or undamaged area.

[0038] The process of associating the coordinates of each area with the material property matching results to generate a dynamic ratio mapping table is as follows: Divide the concrete surface into high, medium and low roughness areas through surface roughness detection (such as laser scanning), and assign different color identifications (red - high, yellow - medium, green - low); extract the spatial coordinate range of each area, establish an association between the coordinate range of each area and the corresponding material parameters (viscosity, carbon nanotube / microcapsule ratio, etc.) to form a static mapping table; monitor the repair process data in real time and dynamically adjust and correct the material parameters (such as adjusting viscosity or additive ratio); write the corrected material parameters into the database in real time to form a dynamically updated ratio mapping table.

[0039] Call the corresponding type of epoxy resin according to the dynamic ratio mapping table, and divide and pack carbon nanotubes and microcapsules according to the allocated ratio; put component A (epoxy resin + carbon nanotubes) and component B (curing agent + microcapsules) into the mixing tank according to a weight ratio of 4:1, and set the mixing parameters to obtain a batch of mixed primer; scan the label of the primer batch to confirm that the area number is consistent with the ratio table; bind the path coordinates in the G code with the roughness level (high / medium / low) in the ratio table to generate a coordinate - nozzle parameter instruction set.

[0040] Further explanation, the specific process of generating the coordinate - nozzle parameter instruction set is as follows: High-roughness area: match a pressure of 0.3 MPa and a flow rate of 0.25 mL / s; medium-roughness area: match a pressure of 0.25 MPa and a flow rate of 0.2 mL / s; low-roughness area: match a pressure of 0.2 MPa and a flow rate of 0.15 mL / s.

[0041] When the robot arm enters the high-roughness area according to the path, it automatically switches to the high-viscosity primer storage tank and moves at a uniform speed along the spiral path. The nozzle sprays the primer according to the corresponding nozzle parameters, and the infrared temperature measurement module monitors the temperature of the glue layer synchronously. According to the captured temperature gradient of the glue layer, the penetration depth is inverted and the penetration depth threshold is set. When the penetration depth is less than the penetration depth threshold, the corresponding coordinate is marked and the robot arm is paused, and it retreats to the previous path, the nozzle is pressurized for re-coating, and the penetration depth is remeasured until the standard is met.

[0042] After the primer is applied, import the finite element optimization path file, and generate the expected strain distribution through simulation; lay the contour coordinates for the carbon fiber cloth and generate the laser cutting template file; select the corresponding carbon fiber cloth batch according to the base roughness level (high / medium / low); fix the selected carbon fiber cloth coil on the laser platform, align the calibration origin with the laser cutting template benchmark, cut according to the path of the laser cutting template, and use the ion fan to process the cutting edge to eliminate static electricity; the six-axis robot arm absorbs the cut carbon fiber cloth piece, arranges the copper electrode belt along the main reinforcement direction of the carbon fiber cloth, uses a multimeter to detect the loop resistance, and performs a continuity test; after the continuity test is completed, load the initial current parameters, and The initial current parameters are dynamically adjusted according to the historical impact load peak value; the dynamically adjusted current parameters are associated with the electrode grid coordinates to generate time-sharing power-on instructions, which are activated in the order of center → edge → transition zone to avoid thermal stress superposition and generate a temperature-strain timing curve; the temperature-strain timing curve is superimposed with the finite element expected strain distribution to identify the hysteresis area, and the current parameters are corrected to generate a strain compensation calibration report; the pressure parameters in the strain compensation calibration report are read to extract the coordinates of the repair area; the pressure parameters are bound to the finite element optimized path coordinates and converted into CNC instructions executable by the robot arm, which automatically pressurizes the repair area and generates a real-time pressure distribution thermal map.

[0043] S5. Extract the coordinate regions with pressure ≥ 1.5 MPa from the real-time pressure distribution thermal map and mark them as high-stress regions; based on the stress vector data in the finite element path, determine the principal stress directions of each high-stress region; arrange fiber optic sensors at 10 cm intervals along the principal stress direction and encrypt the cross regions to 5 cm to generate a fiber optic sensing network layout diagram; the six-axis robotic arm cuts micro-channels along the coordinates of the layout diagram and synchronously vacuum-adsorbs debris; embed the fiber optic sensors into the channels and fix them by dotting ultraviolet curing glue; the grating demodulator measures the initial central wavelength and eliminates sensors with abnormal deviations; connect every 10 FBG sensors in series as a collection channel, and a total of 512 channels are connected to the distributed demodulator; dynamically adjust the LoRa spreading factor, bandwidth, and coding rate according to the signal strength to obtain a synchronous multi-channel acquisition mechanism; load the temperature baseline of the carbon fiber cloth layer and assign an initial temperature reference value to each channel; according to the temperature sensitivity formula, calculate the wavelength shift caused by temperature channel by channel; subtract the temperature component from the total wavelength shift to obtain the strain component and convert it into a strain value according to the strain sensitivity; split the wavelength shift of each channel into a two-column matrix including the strain value and temperature to obtain the decoupled strain-temperature data.

[0044] Further explanation, splitting the wavelength shift of each channel into a two-column matrix including the strain value and temperature means calibrating the temperature sensitivity coefficient through experiments, that is, the wavelength shift corresponding to a unit temperature change, monitoring the temperature change using an internal temperature sensor or an external thermocouple, calculating the wavelength shift caused by temperature, and after subtracting the wavelength shift caused by temperature, the remaining wavelength shift is the strain component, which is converted into a strain value through the strain sensitivity coefficient.

[0045] Extract the decoupled temperature data and calculate the average ambient temperature every 10 minutes; generate a dynamic compensation coefficient according to the deviation between the average ambient temperature and the baseline temperature; multiply the strain value by the dynamic compensation coefficient to obtain a temperature-compensated strain matrix and eliminate the false strain caused by thermal expansion.

[0046] Match the coordinates of the real-time strain values in the temperature-compensated strain matrix with the grid nodes of the three-dimensional damage map to generate a unified spatio-temporal index; take three-dimensional voxels as units and fill in the strain values in time series to form a four-dimensional data cube; map the initial damage depth and expansion rate in the three-dimensional damage map to the corresponding voxels in the four-dimensional data cube as the initial state of time; divide the time window length to segment the four-dimensional data cube, calculate the strain change increment within the window for each voxel, and convert the strain change increment into a damage increment; perform DBSCAN clustering on the damage increment within each voxel to generate a damage evolution map.

[0047] Based on the spatio-temporal four-dimensional data cube of the damage evolution atlas, a spatio-temporal convolutional neural network is used to extract damage evolution features. Through transfer learning, the finite element simulation stress field data is fused to construct a damage development prediction model. The generative adversarial network is used to simulate accelerated aging scenarios under different environmental temperature and humidity conditions to generate a multi-scenario damage propagation path library. Combining with the conductivity degradation curve of carbon fiber cloth, a non-linear mapping relationship between the fiber-matrix interface bond slip and the strain field is established. Through digital twin technology, a full-life cycle health monitoring platform for the reinforced structure is constructed to output the predicted value of the remaining life of the structure and the priority maintenance decision in real time, and generate a structural monitoring and evaluation report.

[0048] Further explanation, constructing a full-life cycle health monitoring platform for the reinforced structure through digital twin technology means integrating real-time monitoring data (such as fiber optic sensor strain, temperature), three-dimensional damage atlas, finite element simulation results, etc., constructing a digital virtual model of the physical structure, and embedding a trained damage prediction model (such as spatio-temporal convolutional neural network). The generative adversarial network is used to simulate multi-environment aging scenarios, and combining with the performance degradation law of carbon fiber cloth, a non-linear mapping relationship between damage evolution and strain field is established.

[0049] This embodiment also provides a computer device, which is applicable to the detection and analysis method of carbon fiber cloth reinforced damaged concrete structures, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the detection and analysis method of carbon fiber cloth reinforced damaged concrete structures proposed in the above embodiment.

[0050] This computer device can be a terminal. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0051] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, a magnetic disk or an optical disc.

[0052] In summary, through the coincidence of the focal points of pulsed laser and continuous laser and the optoelectronic synchronous triggering mechanism, the present invention ensures the alignment of the time stamps of the acoustic emission signal and the vibration frequency offset, improves the correlation accuracy between the transient stress wave propagation characteristics and the material dynamic response, and solves the phase distortion problem caused by signal delay in the traditional method; uses bicubic spline interpolation to convert the discrete depth matrix into a three-dimensional damage distribution map, combines the hydraulic impact to simulate the crack propagation behavior under dynamic load, and optimizes the subspace probability distribution through the GF deviation, significantly improving the spatial resolution of the damage depth and the propagation rate; through the fusion of the damage evolution map and the digital twin model, dynamically adjusts the plasma etching parameters and the electrical conductivity of the carbon fiber cloth, improves the tensile strength of the repair area, and improves the strain decoupling accuracy of the fiber optic sensing network, realizing the full-life cycle health monitoring.

[0053] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth, characterized in that: include, Pulse laser and continuous laser are used to perform grid scanning on the concrete surface, and the amplitude attenuation value and frequency offset data are collected synchronously, and fused and processed to generate a three-dimensional damage distribution map; A high-risk point set is extracted from the three-dimensional damage distribution map, and weighted K-means clustering is performed to construct the Voronoi subspace. Through Monte Carlo simulation, the failure probability is calculated and the emergency repair block set is divided; The repair blocks are prioritized to generate a processing queue, and the spiral path of the plasma spray gun is planned using the TSP algorithm. The surface roughness distribution map is generated by combining laser confocal and white light interferometry detection. According to the surface roughness distribution map, the dynamic ratio scheme of epoxy resin and carbon fiber is matched by partition, and the pressure thermodynamic map is generated through finite element simulation and electrode time-sharing power supply; Deploy fiber optic sensing networks to decouple strain-temperature data, build and train damage development prediction models, predict remaining life, and generate structural monitoring assessment reports.

2. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The generating of the three-dimensional damage distribution map comprises the following steps: Calibrate the energy parameters of pulsed lasers and continuous lasers, and plan the scanning path; Synchronously stimulate the acoustic emission signal and measure the frequency offset, align the amplitude attenuation value and the frequency offset data stream, and generate the amplitude matrix and frequency matrix after removing the abnormal values; Based on the initial amplitude sequence of the impact center point, a discrete depth matrix is ​​generated by integrating the logarithmic decay rate. The three-dimensional damage distribution map is output through bicubic spline interpolation and the high-risk areas are marked.

3. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The calculation of failure probability and division of the emergency repair block set includes the following steps: A set of high-risk points is extracted from the three-dimensional damage distribution map, weighted based on the damage depth and extension rate, and the Voronoi subspace is generated using weighted K-means clustering. Inject normal distribution random perturbations into the subspace and calculate the cumulative failure probability; The subspace weights and cumulative failure probabilities are integrated to generate a failure weight matrix, and DBSCAN clustering is performed to generate a set of emergency repair blocks sorted by priority.

4. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The generating of the surface roughness distribution map comprises the following steps: The repair block set is expanded according to the priority and the overlapping area is detected, and the spiral path is generated using the TSP algorithm; Match the plasma power and helium-oxygen ratio according to the damage depth of the repaired area, write the G code and adjust the etching parameters in real time; Laser confocal microscope is used to detect the etching depth and make additional etching, and white light interferometer scanning is used to generate a surface roughness distribution map.

5. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The generating of the pressure thermodynamic map comprises the following steps: The area is divided according to the roughness distribution map, and the viscosity of epoxy resin and the proportion of carbon nanotubes are matched to generate a dynamic ratio scheme; After spraying the primer, the coating pressure is adjusted based on the penetration depth feedback, and the expected strain distribution is generated by simulation; The carbon fiber cloth is cut and the electrodes are arranged. The temperature-strain curve is generated by power-on at different times, and the pressure thermodynamic diagram is output by superimposing finite element data.

6. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The construction of the damage development prediction model comprises the following steps: The coordinates of high stress areas are extracted from the pressure thermodynamic map, and the optical fiber sensing network is deployed along the principal stress direction; Decouple the strain and temperature data of the optical fiber sensor, eliminate the interference of thermal expansion through dynamic compensation, and generate the temperature-compensated strain matrix; The strain matrix and the three-dimensional damage map are aligned in time and space to construct a damage development prediction model.

7. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: Training the injury progression prediction model includes the following steps: The damage development prediction model is segmented into time windows, and the strain increment is calculated voxel by voxel and converted into damage increment. The spatiotemporal convolutional neural network is used to extract the damage evolution characteristics, and the damage prediction model is trained with finite element simulation data. The generative adversarial network is used to simulate multi-scenario aging paths, and the mapping relationship between carbon fiber cloth performance degradation and damage extension is established.

8. The method for detecting and analyzing damaged concrete structures reinforced with carbon fiber cloth according to claim 1, characterized in that: The generating of the structural monitoring and assessment report refers to constructing a full life cycle health monitoring platform for the reinforced structure through digital twin technology based on a trained damage development prediction model, outputting the predicted value of the remaining life of the structure and the priority maintenance decision in real time, and generating a structural monitoring and assessment report.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth as described in any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for detecting and analyzing a damaged concrete structure reinforced with carbon fiber cloth as claimed in any one of claims 1 to 8 are implemented.

Images