Health assessment methods for long-distance underground culvert structures

Abstract

This invention discloses a method for structural health assessment of long-distance underground culverts, comprising: fusing acoustic and optical detection data to identify structural defects; performing underwater-to-dry conversion and correction of strength values ​​based on underwater rebound and core sampling test data to determine the corrected structural strength and material degradation parameters; combining defect, degradation, and historical monitoring data to perform structural evolution modeling and abrupt change risk identification, obtaining the structure's evolution characteristics and abrupt change risks; and comprehensively assessing the corrected structural strength, evolution characteristics, and abrupt change risks to generate structural health assessment and risk warning results. This invention enables accurate, dynamic, and comprehensive assessment of the culvert's health status, improving the reliability and early warning capability of the assessment results.

Description

Technical Field

[0001] This invention belongs to the field of civil engineering and structural health monitoring, and in particular, it is a method for assessing the structural health of long-distance underground culverts. Background Technology

[0002] Long-distance underground culverts are typically buried deep underground, traversing complex geological and environmental units such as rivers and highways, and operating under harsh conditions of full-water, high-velocity, and sand-laden water for extended periods. Accumulated hydraulic erosion, material aging, and geological changes inevitably lead to problems such as concrete cracking, material deterioration, joint sealing failure, soil seepage around the structure, and even piping. Sudden structural failure not only halts the project but may also trigger secondary risks to the surrounding environment. Therefore, conducting systematic, accurate, and forward-looking health assessments of long-distance underground culverts, and promptly identifying and issuing early warnings of potential risks, is of theoretical significance and engineering application value for ensuring their safe and efficient operation throughout their entire life cycle.

[0003] Currently, technologies for the detection and evaluation of long-distance underground culverts have made significant progress. In terms of detection methods, traditional manual diving exploration is gradually being supplemented or replaced by underwater robots (ROVs) equipped with various sensors. ROVs are typically equipped with high-definition optical cameras, forward-looking sonar, or multibeam echo sounders, enabling preliminary investigation of siltation and concrete surface defects (such as cracks and spalling) within the culvert without emptying it. Regarding structural performance evaluation, in-situ non-destructive testing (NDT) techniques are being applied, such as using the rebound method to sample and test concrete strength in areas with suitable conditions. Simultaneously, fixed monitoring sensors, such as piezometers, strain gauges, and displacement gauges, are deployed at key locations within the culvert to obtain long-term monitoring data on critical structural physical quantities. Furthermore, in terms of analytical methods, numerical simulation techniques based on the finite element method (FEM) are widely used to analyze the seepage and stress field distributions of culverts under design loads and boundary conditions, providing a theoretical basis for understanding the fundamental mechanical behavior of the structure. The application of these technologies and methods forms the foundation of current culvert health assessment work.

[0004] However, while existing technologies have achieved some degree of detection of culverts, they still have significant limitations in the deep fusion of multi-source heterogeneous data and the spatiotemporal evolution assessment of structural health status. Specifically, at the data level, the lack of a unified physical model for spatiotemporal alignment, underwater environmental interference correction, and defect impact quantification of detection data from different sources (such as acoustic, optical, and mechanical data) leads to low accuracy and unclear physical meaning in the fusion results, creating information silos and inaccurate assessments. At the assessment framework level, existing methods typically stop at a static, discrete description of the current state. They cannot reveal the dynamic evolution of structural health under multi-field coupling through a unified evolution function, nor can they reliably reconstruct sparse monitoring data into a continuous state across the entire field based on physical constraints. This dual limitation of data and framework makes it difficult for existing technologies to provide accurate and timely early warnings of potential risks to long-distance culverts. Summary of the Invention

[0005] The purpose of this invention is to provide a method for assessing the structural health of long-distance underground culverts, in order to solve the aforementioned problems existing in the prior art.

[0006] Technical solutions, including methods for structural health assessment of long-distance underground culverts, comprising:

[0007] Based on acoustic and optical detection data, the system performs fusion processing to identify and output structural defect data that includes soil type, defect type, and geometric parameters.

[0008] Based on pre-stored underwater rebound and core sampling test data, combined with structural defect data, underwater-to-dry conversion and correction of strength values ​​are performed to determine the corrected structural strength and material degradation parameters.

[0009] By combining structural defect data, material degradation parameters, and pre-stored historical monitoring data, structural evolution modeling and mutation risk identification are performed to obtain the evolution characteristics and mutation risks of the structure.

[0010] By comprehensively assessing the structural strength, evolutionary characteristics, and mutation risks after correction, a full-field state assessment is conducted to generate structural health assessment and risk warning results.

[0011] Beneficial effects: This invention enables accurate, dynamic, and comprehensive assessment of the health status of culverts, improving the reliability of assessment results and early warning capabilities. Attached Figure Description

[0012] Figure 1 A flowchart illustrating the steps of a method for assessing the structural health of long-distance underground culverts provided in this application embodiment.

[0013] Figure 2 A flowchart illustrating the steps for obtaining the evolutionary characteristics and mutation risks of a structure, as provided in this application embodiment.

[0014] Figure 3 A flowchart illustrating the steps for generating an evolution time series provided in this application embodiment.

[0015] Figure 4 A flowchart illustrating the steps for dynamically calibrating evolution weight coefficients in an embodiment of this application. Detailed Implementation

[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0018] The study revealed that data silos and inaccuracies at the data level lead to insufficient accuracy and completeness in the assessment criteria. Existing methods obtain multi-source data that is physically fragmented and uncalibrated. The physical foundation for acoustic-optical data fusion is weak. Due to physical separation and platform movement, the optical and acoustic sensors on ROVs exhibit inherent spatiotemporal biases in their data, making true feature fusion impossible with simple image overlay. More importantly, the attenuation and scattering patterns of acoustic and optical signals differ significantly between different structural zones (such as the arch and sidewalls) due to varying materials and water flow conditions. Fixed fusion strategies cannot adapt to this spatial heterogeneity, limiting defect identification accuracy. The interference of the underwater environment on in-situ detection results is not accurately quantified. For example, in underwater rebound testing of concrete strength, the rebound value is simultaneously affected by multiple physical effects such as water pressure, water temperature, water flow erosion, and water-medium coupling. Directly using land-based conversion formulas can lead to serious misjudgments of structural strength. Existing methods lack physical models for decoupling and comprehensively compensating for these interfering factors. The geometric information of defects is disconnected from their mechanical effects. The correlation between the location, size, and other geometric information of detected defects such as cracks and spalling and their actual weakening effect on the strength of surrounding materials is ambiguous. Assessments often only allow for qualitative descriptions or rough reductions, failing to quantitatively and continuously reflect the spatial distribution of the defect's impact, leading to inaccurate local risk assessments.

[0019] Furthermore, the static and discrete nature of the assessment framework limits the globality and forward-looking nature of the assessment results. Existing assessment models are often statically pieced together rather than forming an organic whole. There is a lack of a unified spatiotemporal evolution assessment framework. Assessments typically stop at generating a current defect distribution map or intensity distribution map, failing to answer the core question of how the overall health of the structure has evolved from the past to the present. For multi-physics processes such as seepage, stress, and material degradation, existing methods struggle to unify them into an evolution function that can quantify cumulative damage, thus failing to reveal the dynamic evolution of the structural health state. The assessment results are spatially discrete and lack physical meaning. Whether through in-situ detection or sensor monitoring, the obtained data is essentially discrete point or line information. How to scientifically reconstruct the entire culvert's global state distribution from this sparse data is a key challenge. Traditional spatial interpolation methods only consider geometric distances, ignoring the physical laws of evolutionary state propagation in the structural medium (such as diffusion and propagation), leading to reconstructed global state maps that may deviate significantly from the actual physical processes, failing to accurately identify concentrated risk areas and propagation paths.

[0020] like Figure 1 As shown, a method for structural health assessment of long-distance underground culverts is proposed, including the following steps:

[0021] Based on acoustic and optical detection data, the system performs fusion processing to identify and output structural defect data that includes soil type, defect type, and geometric parameters.

[0022] In this embodiment, acoustic detection data includes time difference, amplitude, and incident angle information of multibeam and side-scan echoes, ultrasonic pulse-echo data, acoustic tomography data, and backscattering or scattering cross-section data used to assess the scattering intensity of the substrate and defects at different frequencies. Optical detection data includes high-resolution visible light images and video frames, lidar point clouds, multispectral or hyperspectral images, and infrared thermal imaging data. This step is the starting point for data-driven analysis, aiming to overcome the limitations of a single data source in complex underwater environments (such as high-turbidity water bodies) by integrating the advantages of different sensors, thereby accurately identifying and locating apparent defects in culverts, such as cracks, spalling, and leakage. Structural defect data provides crucial input information for subsequent strength correction and evolution analysis.

[0023] Based on pre-stored underwater rebound and core sampling test data, combined with structural defect data, underwater-to-dry conversion and correction of strength values ​​are performed to determine the corrected structural strength and material degradation parameters.

[0024] In this embodiment, this step aims to obtain the true performance parameters of the structural materials. Considering that the culvert is in a long-term underwater environment, conventional dry-ground testing methods are no longer applicable. Therefore, a physical model can be constructed to quantify and compensate for the influence of the underwater environment (such as water pressure and temperature) on the rebound test, achieving underwater-to-dry-ground strength conversion. Using structural defect data, the strength values ​​around the defects are locally corrected, ultimately obtaining a strength distribution map that reflects the true state of the structure, and the material degradation parameters are calculated based on the strength distribution map.

[0025] By combining structural defect data, material degradation parameters, and pre-stored historical monitoring data, structural evolution modeling and mutation risk identification are performed to obtain the evolutionary characteristics and mutation risks of the structure.

[0026] In this embodiment, the purpose of this step is to reveal the dynamic patterns of structural health changes over time and the potential risks of sudden changes. Specifically, by integrating multi-source information, an evolutionary model that reflects the coupling effects of multiple physical fields such as permeation and stress is constructed. By solving this model, the evolutionary characteristics that quantify the degree of cumulative structural damage can be obtained, and by performing differential analysis on the evolutionary process, the risk transition moments that may lead to structural failure, i.e., the risks of sudden changes, can be identified.

[0027] By comprehensively assessing the structural strength, evolutionary characteristics, and mutation risks after correction, a full-field state assessment is conducted to generate structural health assessment and risk warning results.

[0028] In this embodiment, the evaluation results from multiple dimensions obtained in the preceding steps are integrated to form a global and unified health status assessment. Sparse data reconstruction technology can be used to expand discrete monitoring point information into a continuous state field covering the entire culvert. Based on this, a nonlinear fusion model is employed to comprehensively calculate structural strength representing current performance, evolutionary characteristics representing historical evolution, and mutation risk representing future risks. Finally, the evaluation results are presented intuitively in the form of a comprehensive health index distribution map, and a tiered early warning scheme is provided.

[0029] According to one aspect of this application, before performing fusion processing to identify and output structural defect data, a step of dynamic spatiotemporal alignment of acoustic and optical data is included. This step aims to address differences between acoustic and optical sensors in terms of physical location, propagation medium velocity, and dynamic changes in the motion platform, ensuring that the data basis for subsequent fusion processing is spatiotemporally consistent. Specifically, this alignment step includes:

[0030] The fundamental propagation time difference is calculated based on the difference in the propagation speed of sound and light in water and the spatial distance between sensors.

[0031] In this embodiment, the spatial distance d between the sensors can be obtained by reading the sensor layout file of the underwater robot (ROV) system, specifically the center coordinates (x, y) of the acoustic sensors. s y s , z s ) and optical camera center coordinates (x c y c , z c The Euclidean distance between x and y is d = sqrt[(x) s -x c ) 2 +(y s -y c ) 2 +(z s -z c ) 2 The speed of sound v in water. s and the speed of light v c It can be calculated using empirical formulas based on real-time monitored environmental parameters such as water temperature and depth. For example, the speed of sound v s Can be derived from v s =1449.2+4.6T-0.055T 2 +0.00029T 3 The calculation is based on +(1.34-0.01T)(S-35)+0.016h, where T is water temperature, S is salinity, and h is water depth. The basic propagation time difference Δt is also calculated. base Then it is Δt base =d(1 / v s -1 / v c ).

[0032] By taking into account at least one disturbance effect introduced by the dynamic changes of the detection equipment, a dynamic time difference correction term is obtained.

[0033] In this embodiment, the disturbance effect preferably includes the motion velocity effect of the underwater robot or the beam spread effect of the acoustic sensor. For example, the correction term Δt introduced by the motion velocity effect. motion It can be represented as Δt motion =d·v ROV / (v s ·v c ), where v ROV This is the real-time velocity vector of the ROV. The correction term Δt introduced by the beam spread effect... spread It can be represented as Δt spread =d·tan(θ s / 2) / v s , where θ s The beam angle of the acoustic sensor.

[0034] The basic propagation time difference and the dynamic time difference correction term are superimposed to generate the comprehensive time difference compensation amount Δt. total For example, Δt total =Δt base +Δt motion +Δt spread .

[0035] Based on the comprehensive time difference compensation, the acquisition time axis of acoustic data and optical detection data is shifted and aligned.

[0036] Specifically, the timestamps of the acoustic data can be shifted by t s_new =t s_old +Δt total Furthermore, the temporally overlapping intervals between the translated data and the optical data are resampled to unify the temporal resolution. Where t s_old t is the timestamp of the original acoustic data acquisition. s_new This is the new timestamp after comprehensive time difference compensation for the acoustic data timeline.

[0037] This embodiment comprehensively considers the difference in acoustic and optical propagation velocities, the ROV's motion velocity vector, and the acoustic beam diffusion effect to calculate the dynamic time difference compensation, achieving precise spatiotemporal alignment of multi-sensor data on the mobile inspection platform. This solves the problem of defect location deviation caused by ROV acoustic data lag. Dynamic compensation improves defect location accuracy, enabling accurate differentiation of parallel cracks and providing reliable spatial coordinate information for subsequent precise repair.

[0038] According to one aspect of this application, after completing spatiotemporal alignment, a fusion process is performed to identify and output structural defect data, specifically including:

[0039] Acoustic and visual feature maps were extracted from acoustic and optical detection data, respectively.

[0040] For example, short-time Fourier transform is performed on time-aligned acoustic data to extract frequency domain features and form an acoustic feature map; edge detection and texture analysis are performed on time-aligned optical data to obtain a visual feature map.

[0041] Based on the structural drawings of the culvert, the culvert body is divided into multiple structural zones with different physical properties in space.

[0042] This division is intended to enable adaptive adjustment of subsequent fusion parameters, as the materials and water flow environment in different zones have varying effects on acoustic and optical signals. For example, based on structural mechanical characteristics, the cross-section of a culvert can be divided into a top zone, a sidewall zone, and a bottom zone.

[0043] For each structural partition, specific fusion parameters are adaptively calculated and generated based on its unique material and water flow environment.

[0044] This embodiment is key to achieving intelligent fusion. Specifically, it includes: for each structural partition, extracting partition state parameters characterizing its physical essence from the corresponding design data and environmental monitoring data. This set of parameters includes at least two of the following: material properties represented by the concrete thickness or steel reinforcement distribution of the partition; and hydrodynamic characteristics represented by the water flow velocity or Reynolds number of the partition. The partition state parameters are then substituted into a preset physical model for calculation, generating unique partition-specific fusion parameters for that partition, used to weight the acoustic and visual feature maps respectively. For example, an acoustic attenuation coefficient model α can be established. i =α0×exp(-β a ×h i )×(1+0.1×ρ i )×(1+0.05×Re i ) and optical scattering coefficient model β i =β0×(1+γ s ×TSS i )×(1-0.2×h i / h max ), and perform normalization processing, where h i h is the thickness of the concrete. max ρ is the maximum thickness of the concrete. i Re is the reinforcement ratio. i For Reynolds number, TSS i For suspended solids concentration, α0, β a ,β0,γ s These are the model coefficients.

[0045] Within each structural partition, the acoustic and visual feature maps are weighted and fused using partition-specific fusion parameters to generate a fused feature map. Defects are then identified based on this fused feature map, ultimately outputting structural defect data. The fusion formula can be expressed as F... i =α i ×A i +β i ×V i A i and V i These are the normalized acoustic and visual features, respectively.

[0046] This embodiment addresses the problem of significant differences in detection conditions across different parts of a culvert by adaptively calculating the acoustic-optical fusion weights based on physical parameters such as concrete thickness, reinforcement distribution, and Reynolds number of the water flow in each zone. In the low-velocity top region, the optical weight is increased to obtain clear crack images; in the high-sediment-laden bottom region, the acoustic weight is increased to penetrate turbid media; and a balanced fusion strategy is employed in the transition areas of the sidewalls. This improves the overall defect detection rate of long culverts, providing reliable data support for preventative maintenance.

[0047] Furthermore, after weighted fusion, the process includes steps for verifying and correcting the fusion results to improve their robustness. Specifically:

[0048] For a specific structural partition, the cross-modal mutual information (MI) between the acoustic and visual feature maps corresponding to that partition is calculated, and this MI is used as a quantitative indicator to measure the consistency between the two modalities. i V i The larger the MI(A) value, the higher the consistency of information provided by the two data sources. Optionally, MI(A) can be further defined. i V i = ∑p(a,v)log[p(a,v) / (p(a)p(v))], where p(a,v) is the joint probability that the acoustic feature value falls in quantization level a and the visual feature value falls in quantization level v, p(a) is the marginal probability that the acoustic feature falls in quantization level a, and p(v) is the marginal probability that the visual feature falls in quantization level v.

[0049] When the cross-modal mutual information is lower than a preset confidence threshold (e.g., MI < 0.3), a modal information conflict is determined to exist in that region. This usually means that one of the sensors in that region has been severely interfered with or has failed.

[0050] For partitions with conflicting modal information, the fusion strategy is automatically switched from weighted fusion to a selection-based strategy. This involves choosing the feature with higher confidence (e.g., better signal quality or feature strength) from both acoustic and visual features, and using this higher-confidence feature as the final fusion result for that partition. This process corrects the fused feature map. This effectively avoids contaminating the final fusion result due to poor quality data from a single modality, thus improving the accuracy of defect identification.

[0051] This embodiment ensures the reliability of the fusion results. When turbulent water flow at culvert bends causes blurred optical images but clear acoustic signals, the system intelligently selects the acoustic data; while in the straight sections with stable water flow, it fully utilizes the complementary advantages of the two modes. This reduces the false alarm rate of fusion detection while maintaining high sensitivity, making it particularly suitable for complex environments with frequently changing detection conditions in long-distance culverts.

[0052] According to one aspect of this application, the steps for underwater-dry land conversion and correction of intensity values ​​specifically include:

[0053] Based on core sampling test data, an underwater-dry land integrated conversion coefficient was constructed and calibrated.

[0054] The underwater-dry integrated conversion coefficient in this embodiment aims to compensate for the influence of the underwater environment, and is intended to systematically eliminate the interference of the underwater environment on the rebound detection results. Specifically, it includes:

[0055] For the physical effects of the underwater environment, independent environmental correction factors are quantified and determined. This set of environmental correction factors includes at least two of the following, and preferably all four:

[0056] The pressure correction factor k, which characterizes the effect of hydrostatic pressure on the rebound value, is p Due to hydrostatic pressure p w This will generate pre-stress on the concrete surface, increasing the energy dissipation of rebound impacts, therefore correction is necessary. Wherein, p w =ρ w ×g×h,ρ w The density of water (e.g., 1000 kg / m³) 3 g is the acceleration due to gravity (approximately 9.8 m / s²). 2 h represents the water depth at the detection point. The pressure correction coefficient can be obtained from model k. p =1 / (1+λ p ×p w ) calculate, where λ p The pressure influence coefficient (e.g., λ) calibrated through experiments. p =2×10 -6 / Pa).

[0057] Temperature correction factor k, which characterizes the effect of water temperature on the elastic modulus of concrete. T Water temperature T affects the elastic modulus of concrete, and thus the rebound value. This correction factor can be derived from k. T =[1+α T ×(T-20)]×k age Calculate, where α T The temperature sensitivity coefficient of the elastic modulus (e.g., α) T =-0.0004 / ℃), k age An age correction term is added to account for the effects of early hydration reactions, for example, when the concrete age t... age When less than 90 days, k age =0.9+0.1×(t age / 90).

[0058] The scour correction factor k characterizes the effect of water flow scour on the surface hardness of concrete.e Long-term water erosion will wear down the surface of concrete, reducing its hardness. This effect can be illustrated by model k. e =1-β e ×η e Quantization is performed, where η e The surface hardness loss rate can be calculated from the cumulative scouring depth Δ. e Compared with the average particle size d of aggregate agg The ratio (η) e =Δ e / d agg Estimate, Δ e It is also related to the water flow velocity v and the service time t. service Related; β e To flush out influence factors (e.g., β) e =0.15).

[0059] The media coupling coefficient k characterizes the influence of water medium on the energy transfer of rebound impact. c The presence of water increases damping during the rebound impact process, reducing energy transfer efficiency. This effect can be measured by k... c Make corrections, for example, k c It can be calculated from the baseline efficiency ratio, taking into account the effect of water depth, such as k. c =1.31×(1-0.001×h).

[0060] By coupling and integrating all environmental correction factors, an underwater-dry land integrated conversion coefficient is generated.

[0061] Specifically, the integration step includes: multiplying multiple independent environmental correction factors within the group to obtain a preliminary comprehensive coefficient k=k p ×k T ×k e ×k c Furthermore, a second-order interaction correction term Δk, designed to characterize the coupling effect between different physical effects, is introduced to finally correct the preliminary comprehensive coefficient, generating the underwater-dry land comprehensive conversion coefficient k. final Because the physical effects are not completely independent—for example, the coupled effect of high water pressure and low temperature may be greater than the product of their independent effects—this second-order interactive correction term can be expressed as Δk = 0.05 × (k p -1)×(k T -1), and the final comprehensive conversion coefficient is k. final =k×(1+Δk).

[0062] By applying the underwater-dry land integrated conversion coefficient, the underwater rebound is converted into a preliminary dry land equivalent strength value.

[0063] For example, formula f' can be used c =kfinal ×(A×R w -B) The original value of underwater rebound R w Converted to preliminary dry soil equivalent strength value f' c Where A and B are the calibration parameters of the rebound hammer.

[0064] Based on the defect type and geometric parameters, the preliminary dry equivalent strength value is locally corrected to obtain the corrected structural strength. Based on the comparison between the corrected structural strength and the preset design strength, material degradation parameters are generated.

[0065] In this embodiment, the local correction step specifically includes:

[0066] For each defect in the structural defect data, a three-dimensional influence field model I(r) is constructed based on the geometric parameters of the structural defect data, showing that the influence intensity decreases with increasing spatial distance. For example, this model can be expressed as a Gaussian decay function I(r) = I0 × exp(-r 2 / 2σ 2 ), where r is the distance from any point in space to the defect center, I0 is the influence intensity of the defect center (normalized to 1), and σ is the influence radius, the value of which is related to the length l and width w of the defect, such as σ = sqrt(l 2 +w 2 ) / 4.

[0067] Based on the defect type, a basic strength reduction factor is set to characterize the degree of maximum impact of the defect type. For example, for cracks, the reduction factor k is... crack It can be related to the crack width w, k crack =1-0.15×(w / w cr ) 0.5 , where w cr The critical width; for peeling, its reduction factor k spall It can be related to the peeling depth d.

[0068] By integrating the three-dimensional influence field model with the foundation strength reduction coefficient, the spatially varying defect correction coefficient field k is calculated. d (x, y, z). Specifically, for a single defect, k d (x, y, z) = 1 - (1 - k) type )×I(r) / I0, where k type This is the basic strength reduction factor for this defect type. Optionally, when the three-dimensional influence field models of multiple defects overlap in space, the method further includes: performing a probabilistic superposition operation on the defect correction coefficients corresponding to each independent defect within the spatially overlapping region to generate a final correction coefficient k that reflects the coupled influence of multiple defects. totalThe final correction coefficient is then used to correct the intensity values ​​within the overlapping region. This probability superposition can be expressed as k total =1-Π(1-k i ), where k i This is the correction factor caused by the i-th independent defect at this point.

[0069] The defect correction coefficient field is applied to the initial dry-ground equivalent strength value to obtain the corrected structural strength. Structural strength f c =k d ×f' c This completes the strength correction of the defect-affected area.

[0070] This embodiment establishes a high-precision underwater-to-dry strength conversion model. This model considers the decrease in rebound value for every 10-meter increase in water depth, the change in elastic modulus when the water temperature deviates by 20°C, the surface hardness decay caused by long-term water erosion, and the impact of water as a coupling medium on the efficiency of impact energy transfer, thus reducing the error in underwater concrete strength assessment. This allows for accurate strength distribution throughout long-distance culverts without requiring extensive core sampling, improving detection efficiency and reducing costs. By establishing a three-dimensional Gaussian attenuation field model of defect influence and employing probabilistic superposition to process overlapping areas of multiple defects, a refined quantification of the impact of defects on concrete strength is achieved. The model considers the elliptical influence zone along the crack direction, the hemispherical influence zone of spalling, and the synergistic deterioration effect of multiple defects, improving the spatial resolution of strength assessment compared to traditional two-dimensional fixed-range reduction. It systematically eliminates the combined influence of the underwater environment and local defects on strength detection, obtaining more accurate and reliable structural strength assessment results than traditional methods, providing a solid data foundation for subsequent structural health assessments.

[0071] like Figure 2 As shown, according to one aspect of this application, the steps for structural evolution modeling and mutation risk identification to obtain the evolutionary characteristics and mutation risks of a structure include:

[0072] By integrating structural defect data, material degradation parameters, and historical monitoring data, an evolution function coupled with multi-physics fields is constructed and solved to generate an evolution degree time series that quantifies the cumulative evolution degree of the structure.

[0073] In this embodiment, the evolution degree time series Φ(t) is a dimensionless parameter with a value range between [0, 1], where 0 represents a structurally intact state and 1 represents a structurally completely failed state or reaching the final stage of evolution. Its value directly reflects the degree of cumulative damage or health deterioration of the structure.

[0074] Furthermore, such as Figure 3 As shown, the steps for constructing and solving the evolution function of the multiphysics coupling to generate the evolution degree time series specifically include:

[0075] From historical monitoring data, extract at least several physical field indicators including hydraulic gradient i, seepage flow q, and soil strain ε, and calculate the time increment sequence of each of the physical field indicators.

[0076] Specifically, historical monitoring data can come from various sensors deployed on the culvert, such as pressure gauges, flow meters, and strain gauges. For evolutionary analysis, these raw time-series data need to be preprocessed, for example, by standardizing the time resolution to 1 hour and calculating the change within a time step Δt, thus obtaining the hydraulic gradient increment sequence Δi(t) = i(t) - i(t - Δt), where i(t) is the hydraulic gradient at time t, and Δt is the time step between two adjacent monitoring data points; the seepage flow rate change sequence Δq(t) = [q(t) - q(t - Δt)] / A seep Where q(t) is the seepage flow rate at time t, and A seep Let be the seepage area; and let be the strain rate increment sequence Δε(t) = ε*(t) - ε*(t - Δt), where ε*(t) is the soil strain rate at time t, and ε*(t) = dε(t) / dt.

[0077] Based on soil type and material degradation parameters, the evolution weight coefficients corresponding to each physical field index are dynamically calibrated.

[0078] In this embodiment, the contribution of different physical field processes to structural degradation is not constant, but closely related to the specific environment in which the structure exists (such as soil type) and its own state (such as the degree of material degradation). Optionally, as Figure 4As shown, the dynamic calibration of evolution weight coefficients includes: configuring basic weight coefficients for each physical field index based on soil type; extracting structural deterioration degree indices from material deterioration parameters and correcting at least one coefficient value in the basic weight coefficients accordingly to obtain adjusted weights; and performing normalization calculations on the adjusted weights to generate evolution weight coefficients. Specifically, basic weight coefficients are configured for each physical field index based on the soil type included in the structural defect data. For example, based on soil mechanical properties, for sandy soil that is more sensitive to seepage failure, higher basic weights can be configured for hydraulic gradient i and seepage flow q (e.g., α=0.5, β=0.3); while for clay that is more sensitive to deformation, higher basic weights can be configured for soil strain ε (e.g., γ=0.3). Structural deterioration degree indices (e.g., strength loss rate η) are extracted from material deterioration parameters, and at least one coefficient value in the basic weight coefficients is corrected based on these structural deterioration degree indices to obtain adjusted weights. For example, when the strength loss rate η exceeds a certain threshold (e.g., η>0.2), it indicates that the structural bearing capacity has decreased and it is more sensitive to strain. In this case, the weight of the strain term can be increased, such as the corrected weight γ'=γ×(1+0.5η). Normalization calculation is performed on the adjusted weights to make the sum of the coefficient values ​​equal to one, generating the final evolution weight coefficients α', β', γ'.

[0079] The time increment sequence is weighted and fused using evolution weight coefficients to obtain the evolution increment sequence dΦ / dt, which reflects the instantaneous evolution rate of the structure.

[0080] In this embodiment, the instantaneous evolution rate can be expressed as dΦ / dt=α'×Δi(t)+β'×Δq(t)+γ'×Δε(t). Optionally, to eliminate the influence of measurement noise on the calculation results, a Kalman filter can be applied to the evolution increment sequence after calculation.

[0081] Numerical integration is performed on the evolution increment sequence along the time dimension to generate the evolution degree time series.

[0082] Specifically, numerical integration methods such as the trapezoidal integral method can be used to accumulate dΦ / dt step by step from the initial state Φ0 (e.g., Φ0=0), i.e., the evolution degree time series Φ(t)=Φ0+∫(dΦ / dt)dt. Through this step, a curve that continuously reflects the evolution of the culvert structure from its past to its current healthy state can be obtained, providing a basis for subsequent evolution stage division and mutation risk identification.

[0083] This embodiment achieves precise matching between the multiphysics coupling model and actual geological conditions by dynamically adjusting the evolution weight coefficients based on soil type (clay / silt / sand) and material degradation degree. When the material degradation rate exceeds 20%, the system automatically increases the strain weight, paying more attention to the impact of structural deformation; in sandy areas, the hydraulic gradient weight is increased to focus on monitoring the risk of seepage failure. This allows the same evolution model to adapt to the complex and variable geological conditions along the culvert route, improving prediction accuracy compared to fixed-weight methods, especially in terms of evolution assessment accuracy at key locations such as stratigraphic interfaces.

[0084] Differential analysis of the evolution time series identifies potential abrupt change moments that characterize risk transitions.

[0085] The purpose of this embodiment is to capture nonlinear and nonstationary key nodes from continuous evolution curves that indicate qualitative changes in the structural state. Specifically, it identifies potential abrupt change moments characterizing risk transitions, including:

[0086] Solving for the first and second derivative sequences of the evolution degree time series; high-precision numerical methods can be used to solve for the first derivative sequence ΨΦ / Ψt and the second derivative sequence Ψ of the evolution degree time series Φ(t). 2 Φ / Ψt 2 Ψ and Ψ represent the evolution rate and acceleration of the structure, respectively, where Ψ is the partial derivative.

[0087] In this embodiment, to improve calculation accuracy and avoid amplifying numerical errors, the five-point central difference method is preferably used for differentiation. For example, the first derivative can be calculated using [-Φ(t+2Δt)+8Φ(t+Δt)-8Φ(t-Δt)+Φ(t-2Δt)] / (12Δt). Furthermore, to eliminate the influence of noise in the original data on the derivative calculation, the derivative sequence can be smoothed using Savitzky-Golay filtering.

[0088] A time series of mutation sensitivity is constructed by calculating the ratio of the absolute values ​​of the second derivative sequence and the first derivative sequence at each time step.

[0089] Specifically, mutation sensitivity time series S m (t)=|Ψ 2 Φ / Ψt 2 The mutation sensitivity here is a dimensionless indicator that amplifies the change in evolutionary acceleration relative to the current evolutionary rate. Physically, it characterizes the degree to which the evolutionary trend reaches an inflection point or undergoes a drastic change. When the structure is in a stable evolutionary phase, this ratio is small; however, when the structural state is about to undergo or is undergoing a mutation, the evolutionary acceleration changes drastically, causing this ratio to peak.

[0090] Each sensitivity value in the mutation sensitivity time series is compared with a preset threshold, and the moments that exceed the threshold are identified as potential mutation moments.

[0091] Specifically, the threshold can be a fixed value (e.g., S). cr =2.5), or, in a preferred embodiment, an adaptive threshold is used. For example, by calculating the mutation sensitivity time series S m (t) Moving average S* over a sliding time window m and standard deviation σ s The mutation threshold is dynamically adjusted to S. cr =S* m +n×σ s (where n can be 2.5), so that the threshold can adapt to the baseline fluctuations at different evolutionary stages and improve the accuracy of identification.

[0092] This embodiment constructs a mutation sensitivity index by calculating the ratio of the second derivative to the first derivative of the evolution time series, achieving precise capture of nonlinear mutation points during the seepage evolution of culverts. This index exhibits a peak when the evolution rate changes drastically, effectively identifying the critical moments of transition from the seepage stage to the seepage deformation stage, and from seepage failure to the channel expansion stage. Compared to traditional methods relying solely on seepage flow or displacement monitoring, this approach can predict system instability before significant changes in physical quantities occur, providing earlier warnings of major accidents such as culvert collapses and enhancing the safety assurance capabilities of underground water conservancy facilities.

[0093] Furthermore, to verify the authenticity of the identified potential mutation moments and explore their physical causes, the following steps are also included:

[0094] For each identified potential abrupt change, acoustic data segments within a predetermined time window before and after that potential abrupt change are extracted from the acoustic detection data. For example, if at t m If a potential mutation is identified at any time, then [t] is extracted. m -300s, t m Acoustic waveform data within the +300s interval.

[0095] Perform spectral analysis on the acoustic data segment to decode whether there are acoustic anomaly modes associated with a pre-defined specific failure mechanism.

[0096] This embodiment aims to cross-validate abrupt changes inferred from macroscopic physical field data using acoustic emission signals directly related to physical processes such as structural deformation and cracking. Optionally, the identified acoustic anomaly patterns include at least one of the following characteristics: In the high-frequency band, a sudden increase in spectral energy exceeding a predetermined decibel value. For example, in the frequency band above 20 kHz, a sudden increase in energy exceeding 10 dB of background noise; this pattern is typically associated with brittle failure events such as microcrack propagation in materials or tearing of waterproofing materials. In the low-frequency band, a sustained jump in vibration intensity significantly higher than the background noise level. For example, in the frequency band below 100 Hz, vibration energy consistently exceeding the normal level by more than 5 times; this pattern may correspond to overall structural resonance, particle erosion, or abnormal changes in water flow pulsation pressure.

[0097] If an acoustic anomaly pattern exists, the potential mutation moment is upgraded to a verified mutation risk, and the mutation risk is classified according to its cause based on the acoustic anomaly pattern.

[0098] For example, if a sudden increase in high-frequency energy is identified, the risk can be initially classified as structural cracking risk; if a jump in low-frequency vibration is identified, it can be classified as structural instability or scour risk. This not only improves the reliability of risk warnings but also provides direct physical evidence for subsequent response decisions.

[0099] This embodiment verifies the potential mutation time calculated by mathematical mutation theory by temporally correlating it with the acoustic spectrum anomaly patterns before and after that time, thus achieving a dual confirmation mechanism for mutation risk. When an abnormal peak appears in the second derivative of the evolution degree function, the system automatically extracts acoustic data within the corresponding time window for spectrum analysis, identifying high-frequency noise surges (>20kHz, amplitude increase >10dB) caused by waterstop tearing or low-frequency resonances (<100Hz) caused by structural vibration, transforming abstract mathematical warnings into specific physical damage types. This reduces the false alarm rate of mutations in culvert detection and can identify early signs of seepage damage, providing accurate time windows and failure mode information for maintenance decisions.

[0100] Based on the evolution rate time series and the spatial location of structural defect data, an evolution rate distribution map is derived and generated;

[0101] In this embodiment, the evolution rate over time is correlated with the spatial location of structural defects to generate a dynamic spatial evolution field, visually presenting the differences in deterioration trends across different regions of the structure. Specifically, for each defect point x... i Extract its corresponding local evolution rate Φ*(x) i ,t) (that is, the first derivative ΨΦ / Ψt at position x) i The value); will discrete Φ*(x) iMapping Φ*(x, t) to a continuous structure domain: Φ*(x, t) = ∑ i=1 n λ i Φ*(x i ,t); where the weight λ i The spatial covariance function is optimized to ensure that the interpolation results conform to the spatial correlation of the structural materials. The spatiotemporal evolution rate field Φ*(x, t) (3D space + time tensor) is output; a key time slice t is selected. k (e.g., current moment, historical abrupt change point), extract the slice field Φ*(x, t) k Evolution rate distribution maps can be generated through visualization using contour plots or heatmaps.

[0102] The evolutionary time series, potential mutation times, and evolution rate distribution map together constitute the evolutionary characteristics and mutation risks of the structure.

[0103] In this embodiment, temporal evolution, mutation nodes, and spatial rate fields are integrated to form a multi-dimensional risk criterion, supporting structural safety classification decisions. Specifically, based on the evolution degree time series, potential mutation times, and evolution rate distribution map, an evolutionary characteristic report is generated, including the cumulative evolution Φ(t). now ), representing the overall degree of structural degradation; the current rate maxΦ*(x, t) now ), representing the evolution rate of the most dangerous region; mutation risk S m (t now (Sensitivity) represents the probability of a sudden change in a nearby state. A dynamic risk grading model is constructed, defining the risk level R as a weighting function: R = w1·Φ(t) / Φ crit +w2·maxΦ* / Φ* crit +w3·S m (t) / S cr The weight allocation (example) is as follows: w1=0.3, w2=0.4, w3=0.3; Φ crit The maximum allowable degree of evolution (e.g., concrete expansion limit of 0.5%); Φ*crit is the critical rate (e.g., rock mass creep rate > 1 mm / day); S cr This is the sensitivity threshold.

[0104] According to one aspect of this application, since monitoring points are always sparse in actual engineering, it is impossible to directly obtain the health state of every point on the structure, and simple spatial interpolation methods ignore the physical laws of state propagation in the structural medium. Therefore, a full-field state assessment includes a full-field reconstruction stage to generate a continuous state distribution. Specifically, this full-field reconstruction stage includes:

[0105] Evolutionary parameters contained in evolutionary characteristics and mutation risks are set as sparse state monitoring points in the three-dimensional space of the culvert.

[0106] In this embodiment, the evolutionary parameters refer to the degree of evolution Φ and the evolution rate ΨΦ / Ψt at each key location. The selection of these key locations can be further optimized; for example, priority can be given to selecting locations in regions where the evolution rate gradient changes drastically or where mutation risks have been identified.

[0107] Construct physical constraint equations to describe the spatiotemporal evolution of the propagation of the evolutionary state in the structural medium.

[0108] This embodiment incorporates engineering experience and physical mechanisms into the spatial interpolation process. In one specific implementation, the spatiotemporal evolution physical constraint equation is a partial differential equation. This equation includes at least a time derivative term describing the change of the evolutionary state over time, and a spatial derivative term describing the spatial propagation and diffusion of the evolutionary state. For example, an evolutionary propagation equation of the following form can be constructed: ▽ 2 Φ-(1 / c 2 )Ψ 2 Φ / Ψt 2 =S(x, y, z); where Φ is the degree of evolution of the spatial position (x, y, z) at time t, and is the unknown quantity to be solved; ▽ 2 For the Laplace operator, represents the second derivative term in space (Ψ). 2 / Ψx 2 +Ψ 2 / Ψy 2 +Ψ 2 / Ψz 2 This term describes the spatial diffusion or propagation trend of the evolutionary state; c is the evolution propagation speed, a physical parameter characterizing the speed at which the deterioration or damage effect propagates in the structural medium. Its value can be calibrated according to different medium types (such as sand, silt, concrete). For example, c can be taken as 2 m / day in sandy areas and 0.5 m / day in clay areas; Ψ 2 Φ / Ψt 2 The second derivative term describes the inertia of the evolutionary state as it changes over time; S(x, y, z) is the source term, representing the endogenous growth rate of the evolutionary state at the spatial location (x, y, z), which can be expressed as the degree of evolution Φ at that point itself. local It is determined that S(x, y, z) = κ × Φ local ×(1-Φ local ), where κ is the growth rate coefficient.

[0109] Using data from sparse state monitoring points as the solution boundary or initial condition, the physical constraint equations are numerically solved to generate a full-field evolution distribution map that continuously covers the entire culvert.

[0110] Specifically, the partial differential equation can be solved using the finite element method or the finite difference method. A three-dimensional mesh is created for the culvert; preferably, adaptive mesh refinement is performed in regions with large evolution gradients. The evolution degree values ​​of sparse monitoring points are applied to the model as known conditions (e.g., Dirichlet boundary conditions). Through iterative solving, the evolution degree Φ(x, y, z, t) of all mesh nodes within the entire solution domain (i.e., the entire culvert) can be obtained, forming a spatiotemporally continuous full-field evolution degree distribution map.

[0111] This embodiment achieves physical constraint reconstruction from sparse monitoring points to a continuous distribution across the entire field by establishing an evolutionary propagation partial differential equation containing time and spatial derivative terms. The evolutionary propagation rate is determined based on the permeability characteristics of different soil types, ensuring that the reconstruction results not only match measured values ​​at monitoring points but also follow the physical laws of seepage failure throughout the entire spatial domain. Compared to purely statistical methods such as traditional Kriging interpolation, this physical information constraint allows for accurate identification of high-risk areas with drastic evolutionary gradient changes even under sparse distribution, improving spatial resolution and avoiding non-physical oscillations that may occur with statistical interpolation. By reasonably interpolating and extrapolating sparse, discrete monitoring data to the entire structural space through partial differential equations containing physical mechanisms, the resulting full-field state distribution map more realistically reflects the spatial continuity and evolutionary laws of structural health status compared to traditional interpolation methods.

[0112] Furthermore, assessment information from different dimensions (structural strength, evolutionary history, and mutation risk) is scientifically integrated into a single, intuitive comprehensive health index, thereby achieving a final assessment of the overall health status of the structure. Specifically, the overall site status assessment also includes a comprehensive index fusion stage. This fusion stage includes:

[0113] The full-field evolution degree distribution map, the corrected structural strength, and the mutation risk index in the evolution characteristics and mutation risk are normalized to form a standardized multi-source dataset.

[0114] This embodiment describes data preparation work before data fusion, aiming to eliminate differences in units and numerical ranges between different data sources. Specifically, all data (e.g., the full-field evolution distribution map Φ(x, y, z, t), the corrected intensity distribution map f) can be prepared. c (x, y, z), and the mutation risk index P risk Resampling is performed on a unified 3D spatial grid. Data sources are normalized; for example, the intensity values ​​f are... c Convert to normalized intensity f in the range [0, 1] norm =(f c -f min ) / (f max -f min ), where fmin For the minimum strength value, f max This represents the maximum intensity value. The evolution degree and mutation risk indicators themselves are already within the range [0, 1] and can be used directly. The final standardized dataset {1-Φ} is formed. norm f norm ,1-P risk}, where Φ norm For the normalized global evolution degree, 1-Φ norm and 1-P risk These represent the degree of contribution to health.

[0115] Based on the spatial information entropy of each data source in a standardized multi-source dataset, the data fusion weight is dynamically calculated.

[0116] This embodiment aims to objectively and data-drivenly determine the importance of each evaluation dimension. Specifically, the spatial information entropy H of each data source i can be calculated. i =-Σp ij ×log(p ij ), where p ij This represents the probability that the value from the i-th data source falls within the j-th interval. A data source with higher information entropy indicates a richer distribution across the entire structural space and contains more information; therefore, it should be assigned a higher weight. The initial weight can be set to w. i 0 =H i / ΣH k H k Let w1 be the spatial information entropy of the k-th data source. Optionally, the initial weights can be modified according to the reliability of each data source (e.g., reconstruction uncertainty, detection signal-to-noise ratio, etc.) to obtain the final dynamic weight vector [w1, w2, w3].

[0117] By employing data fusion weights and using a nonlinear fusion model that includes interaction terms, a standardized multi-source dataset is calculated to generate a comprehensive health index distribution map, which constitutes the core part of the structural health assessment and risk warning results.

[0118] In this embodiment, a nonlinear model is used because different risk factors often exhibit a coupling amplification effect, which simple linear weighting cannot capture. This nonlinear model can be expressed as H=H linear +H interact ; where the linear part H linear =w1×(1-Φ norm )+w2×f norm +w3×(1-P risk Interactive section H interact This includes product terms of different factors, such as H. interact =λ12 ×(1-Φ norm )×f norm +..., where λ 12 Let be the interaction coefficient between the i-th and j-th factors.

[0119] Furthermore, the nonlinear fusion model also includes a risk avoidance criterion, the implementation of which includes:

[0120] The system checks each data source value in the standardized multi-source dataset point by point to determine if any value falls below a preset risk threshold. For example, the risk threshold can be set to 0.4.

[0121] If it is determined that a value is below the risk threshold (e.g., the normalized strength f at a certain point), norm If the value is only 0.3, the weighted fusion calculation at that point is stopped, and the lowest value at that data point is directly adopted as the final comprehensive health index, thereby amplifying the risk of the weakest link effect.

[0122] In other words, under these circumstances, the comprehensive health index H at this point is directly determined to be 0.3, even if its evolutionary degree and mutation risk indicators are high. This risk avoidance criterion follows the barrel theory, ensuring that any serious deficiency in a single dimension will not be masked by the good performance of other dimensions, making the assessment results more conservative and safe. The final generated comprehensive health index distribution map H(x, y, z) has a value range between [0, 1], which can be color-coded and displayed in three dimensions. Health levels and risk warnings are then assigned based on different index intervals (e.g., H ≥ 0.85 for healthy, 0.70 ≤ H < 0.85 for basically healthy, etc.).

[0123] This embodiment achieves intelligent fusion and risk amplification of multi-source heterogeneous data. Data sources with high information entropy (such as those with drastic changes in evolution degree in the infiltration and damage area) automatically receive higher weights, while the system proactively adopts the most conservative assessment strategy when key indicators show warning signals. This solves the subjectivity problem of traditional expert weighting methods, and at the same time, the risk avoidance criterion ensures that severe local degradation is not masked by normal indicators, thus improving the reliability of culvert health assessment and avoiding structural failure caused by the weakest link in the chain.

[0124] In a specific embodiment, assuming the detection is carried out on a side wall of a long underground culvert, the specific environment and detection parameters of the measuring point are as follows:

[0125] Environmental parameters are: water depth h = 10m, water temperature T = 15°C, water flow velocity v = 1.5m / s (assuming long-term service), and water density ρ. w =1000kg / m 3The rebound tester parameters are: impact energy E0 = 2.207 J, calibration parameters A = 10, B = 100. The original underwater rebound measurement value R was measured at this point. w =17.5. A vertical crack with a width of w=2mm exists 0.3 meters near this measuring point. The component thickness is 0.5m. Calculate the pressure correction factor k. p hydrostatic pressure p w =ρ w ×g×h=1000×9.8×10=98000Pa. Assume the pressure influence coefficient λ. p =2×10 -6 / Pa, then k p =1 / (1+2×10 -6 (×98000)=1 / 1.196≈0.836. Calculate the temperature correction factor k. T Assuming the concrete age is much greater than 90 days, k age =1. Assume the temperature sensitivity coefficient α T =-0.0004 / °C, then k T =[1+(-0.0004)×(15-20)]×1=1+0.002=1.002. For simplicity, the water flow scour correction coefficient k is tentatively set in this embodiment. e and the medium coupling correction factor k c According to the calibration results, k are respectively e =0.98 and k c =1.25. (In practical applications, these two items should also be calculated using the corresponding model). Calculate the preliminary comprehensive coefficient k: k=k p ×k T ×k e ×k c =0.836×1.002×0.98×1.25≈1.025. Calculate the second-order interactive correction term Δk and obtain k. final Δk=0.05×(k p -1)×(k T -1)=0.05×(0.836-1)×(1.002-1)≈-0.0000164. k final =k×(1+Δk)=1.025×(1-0.0000164)≈1.02498. For clarity, we take k here. final =1.025. Calculate the preliminary dry equivalent strength f'. c :f' c =k final ×(A×R w -B) = 1.025 × (10 × 17.5 - 100) = 1.025 × 75 = 76.875 MPa. Calculate the foundation strength reduction factor k.crack Assuming the critical crack width w cr =5mm. k crack =1-0.15×(w / w cr ) 0.5 =1 - 0.15 × (2 / 5) 0.5 ≈1 - 0.15 × 0.632 = 1 - 0.0948 = 0.9052. Calculate the defect correction coefficient k for spatial variation. d The radius of influence of the defect is assumed to be σ = 0.5m. The distance from the measuring point to the center of the defect is r = 0.3m. The influence intensity I(r) = exp(-r) 2 / 2σ 2 )=exp(-0.3 2 / (2×0.5 2 ))=exp(-0.09 / 0.5)=exp(-0.18)≈0.835. Defect correction coefficient k d =1-(1-k crack )×I(r)=1-(1-0.9052)×0.835=1-0.0948×0.835≈1-0.0792=0.9208. Calculate the final corrected structural strength f. c :f c =k d ×f' c =0.9208×76.875≈70.78MPa.

[0126] As can be seen from this embodiment, the original underwater rebound reading of 17.5 may correspond to a relatively high strength value without correction. However, through multi-factor environmental correction (correcting the strength from approximately 75 MPa to 76.875 MPa) and critical defect impact correction (further correcting the strength from 76.875 MPa to 70.78 MPa), a strength value that better reflects the actual health condition of the structure was obtained.

[0127] In another embodiment of this application, a method for assessing the structural health of long-distance underground culverts includes:

[0128] Step 1: Collect raw acoustic data and optical image sequences from the ROV, perform spatiotemporal fusion through propagation time difference compensation and zonal adaptive correction, identify and locate structural defects, and output spatiotemporally aligned fused detection data, defect feature maps, and defect location coordinate sets.

[0129] Specifically, acoustic waveform data (sampling rate ≥ 100kHz) and optical image sequences (frame rate ≥ 30fps) from the sensors mounted on the ROV are read, and data integrity checks and timestamp synchronization are performed to obtain time-stamped acoustic data streams and time-stamped image data streams. The time-stamped acoustic and image data streams are then read to calculate the difference in sound and light propagation speeds in water (sound speed 1500m / s, light speed 2.25×10⁻⁶). 8 (m / s) The time axis is shifted according to the sensor spacing d to obtain time-aligned acoustic and optical data. The culvert structure CAD drawings are read, dividing the culvert into top, sidewall, and bottom zones. Based on the material properties (concrete thickness, reinforcement distribution) and flow conditions of each zone, the acoustic attenuation coefficient α of each zone is calculated. i and optical scattering coefficient β i The partition fusion parameter table is obtained. Time-aligned acoustic data is read, and short-time Fourier transform is performed to extract frequency domain features (0.1-50kHz) to identify defect feature frequencies. Time-aligned optical data is read, and edge detection and texture analysis are performed to obtain acoustic and visual feature maps. The acoustic and visual feature maps, along with the partition fusion parameter table, are read, and a weighted fusion F is applied to each partition. i =α i ×A i +β i ×V i (A) i For acoustic characteristics, V i Visual features are used to perform cross-modal feature verification, resulting in a fused feature map. The fused feature map is then read, and threshold segmentation and connected component analysis are performed to identify cracks (width > 0.2 mm) and leaks (wetted area > 100 cm²). 2 Defects such as spalling (depth > 5mm) are identified, and their 3D coordinates are calculated using ROV location data to obtain a set of defect location coordinates, defect type labels, and defect geometric parameters. The fused feature map and the original acoustic and visual feature maps are then read to calculate the fusion gain index G=SNR. fusion / (SNR acoustic +SNR visual ), where SNR fusion The signal-to-noise ratio (SNR) of the fused acoustic-optical features. acoustic The signal-to-noise ratio (SNR) of the original acoustic signal. visual The signal-to-noise ratio of the original optical image is used to verify the fusion effect (G>1.5 indicates effective fusion), and spatiotemporally aligned fusion detection data and defect feature maps are generated.

[0130] Step 2: Read the original rebound measurement values, core test data, and defect location coordinate set, perform strength calculations using the underwater-dry conversion model, and output the corrected concrete strength distribution map and material degradation parameters.

[0131] Specifically, the raw rebound measurement value R of the underwater rebound tester is read. w (16 readings at each measuring point) Simultaneously record water depth h, water temperature T, and water flow velocity v, discard outliers (exceeding the mean ±3σ), and obtain the effective rebound value set and environmental parameter set. Read the concrete core samples collected by the core drilling machine, conduct laboratory compressive strength tests, and record the core sample compressive strength value f. c By comparing the set of effective rebound values ​​at the same location, a correlation between rebound value and intensity is established to obtain calibration data pairs. The calibration data pairs and environmental parameter sets are then read, and the pressure correction term k is calculated based on the influence of water pressure. p =1-0.002h, calculate the temperature correction term k based on the temperature effect. T =1+0.01(T-20), and the conversion coefficient k is obtained by combining the results. p ×k T ×k0 (k0 is the baseline coefficient 1.15) yields the underwater conversion coefficient. The defect location coordinate set and defect geometric parameters are read. For rebound test points within a 50cm radius of the defect, strength reduction is applied according to the defect type: cracks reduce by 5-15%, spalling by 10-25%. The correction coefficient k is then calculated. d The defect correction coefficient distribution is obtained. The effective rebound value set, underwater conversion coefficient, and defect correction coefficient distribution are read, and the concrete strength f at each measuring point is calculated. c =k×k d ×(10×R w -100), reconstruct the strength distribution between measuring points to obtain the corrected concrete strength distribution map. Read the corrected concrete strength distribution map and the design strength value, and calculate the strength loss rate η=(f design -f current ) / f design Analyze the intensity gradient ▽f to identify areas of concentrated degradation and extract the degradation rate v. d =Δf / Δt (based on historical data) to obtain the material degradation parameters. Where f design f is the design strength value. current Δf represents the currently measured concrete strength value, and Δf represents the change in strength.

[0132] Step 3: Read the defect feature map, material degradation parameters and historical monitoring data, and output the evolution state parameters, mutation risk index and evolution rate distribution through the continuous evolution model and mutation detection algorithm.

[0133] Specifically, the defect feature map is read to identify seepage traces and water flow paths, and the seepage area A is calculated.w and wet perimeter P w The material degradation parameters were read, and the porosity growth rate and permeability coefficient changes were extracted to obtain permeability characteristic parameters and the initial evolution state. Historical monitoring data (including water level, flow rate, strain, etc.) were read, cleaned, and interpolated to a unified time resolution of 1 hour; the hydraulic gradient i and seepage velocity v were calculated. s The soil strain rate ε* is used to obtain multiphysics time-series data. The multiphysics time-series data and seepage characteristic parameters are read, and the instantaneous evolution increment dΦ / dt = α × Δi + β × Δq + γ × Δε (weighting coefficients α = 0.4, β = 0.35, γ = 0.25 are determined according to the soil type) is calculated. Time integration Φ(t) = ∫dΦ / dt × dt is performed to obtain the evolution degree time series. The evolution degree time series is read, and the current evolution stage is determined according to the thresholds: Φ < 0.2 is the seepage stage, 0.2 ≤ Φ < 0.4 is the seepage deformation stage, 0.4 ≤ Φ < 0.6 is the seepage failure stage, 0.6 ≤ Φ < 0.8 is the channel expansion stage, and Φ ≥ 0.8 is the riverbed collapse stage, obtaining the current evolution stage identifier and stage transition time. The evolution degree time series is read, and the first derivative ΨΦ / Ψt and the second derivative Ψ are calculated. 2 Φ / Ψt 2 Calculate the mutation sensitivity S m =|Ψ 2 Φ / Ψt 2 | / |ΨΦ / Ψt|, when S m Points exceeding 2.5 are marked as potential mutation points, yielding mutation sensitivity curves and a set of potential mutation time points. The time series of acoustic feature maps and the set of potential mutation time points are read, and the spectral changes before and after the mutation time are analyzed to identify high-frequency noise spikes (>20kHz, amplitude increase >10dB) and low-frequency vibrations (<100Hz) characteristics, confirming the mutation type and obtaining mutation risk indicators and mutation type labels. Optionally, obtaining the mutation risk indicators and mutation type labels includes: reading the acoustic feature maps and the set of potential mutation time points, and for each mutation time t... m Extract [t] m -300s, t m Acoustic data within a time window of +300s; perform short-time Fourier transform (window length 1024, overlap 50%), calculate power spectral density PSD(f,t); divide frequency bands: low frequency (<100Hz), mid frequency (100Hz-10kHz), high frequency (>10kHz), calculate energy E for each frequency band. low E mid E high The acoustic feature matrix at the abrupt change moment is obtained. The acoustic feature matrix at the abrupt change moment is read, and the statistics of the characteristic time series are calculated: the high-frequency energy surge rate r. high =(E high_max -E high_mean) / E high_mean , where E high_max is the maximum value of the acoustic energy in the high-frequency band, and E high_mean is the average acoustic energy in the high-frequency band; the low-frequency vibration intensity I low =∫E low (f < 50Hz)df; Identify abnormal patterns: water stop tearing (r high > 3 and lasting > 10s), structural vibration (I low > 5 times the mean value), particle erosion (energy increase in 2kHz < f < 5kHz > 10dB); Obtain the acoustic abnormal pattern label and abnormal intensity index. Read the set of potential mutation time points and the acoustic abnormal pattern label, and calculate the time difference Δt lag = t acoustic - t evolution , where t acoustic is the central time point when a certain acoustic abnormal pattern (such as a sudden increase in high-frequency noise or low-frequency vibration) occurs, and t evolution is the time of a certain evolutionary mutation point identified in the mutation sensitivity analysis; Statistically analyze the typical time delays of different abnormal patterns: water stop tearing (-60s ~ +30s), structural vibration (-120s + 180s); Establish the correlation intensity index R ae = exp(-|Δt lag | / τ), where τ = 60s is the characteristic time; Obtain the acoustic-mutation correlation matrix. Read the preliminary judgment of the mutation type, the acoustic abnormal pattern label, and the acoustic-mutation correlation matrix, and perform multi-source information fusion: When R ae > 0.7 and the abnormal pattern matches, confirm the mutation type; Calculate the comprehensive risk index R risk = w1×I m + w2×r high + w3×R ae , where the weights are w1 = 0.3, w2 = 0.4, w3 = 0.3, and I m is the acoustic abnormal intensity index of this mutation event; Risk classification: R risk > 0.8 is high risk, 0.5 - 0.8 is medium risk, < 0.5 is low risk; Output the mutation risk index and the mutation type label. Read the evolution degree time series, the set of defect position coordinates, and the current evolution stage identifier, and calculate the local evolution rate v Φ = dΦ / dt of each defect point, perform spatial interpolation according to the defect spacing and hydraulic connectivity, and obtain the evolution rate distribution and evolution state parameters.

[0134] Step 4. Read the evolution state parameters, the mutation risk index, the evolution rate distribution, and the corrected concrete strength distribution map, and output the full-field health state distribution map, the structural health assessment report, and the graded early warning plan through sparse reconstruction and comprehensive evaluation models.

[0135] Specifically, the evolution rate distribution and defect location coordinate set are read to identify extreme points in the evolution rate and areas of drastic gradient changes, determining the locations of key monitoring points. This ensures that the distance between any two points is less than 50m and covers all high-risk areas, resulting in an optimized monitoring point set. The evolutionary state parameters of the optimized monitoring point set are then read to establish the evolutionary propagation equation ▽. 2 Φ-(1 / c 2 )Ψ 2 Φ / Ψt 2 =S(x, y, z), where c is the evolution propagation speed (0.5-2 m / day) and S is the source term intensity. Finite element analysis is performed to obtain the overall evolution degree distribution Φ(x, y, z, t). Specifically, the process of obtaining the overall evolution degree distribution is as follows: The defect location coordinate set and evolution rate distribution are read; mesh refinement is performed in high-gradient regions (|▽Φ|>0.1 / m), with a minimum element size of 0.5m; sparse meshes are used in slow-evolution regions (|▽Φ|<0.01 / m), with a maximum element size of 5m; Delaunay triangulation is used to generate an unstructured mesh, ensuring a mesh quality index >0.7; an adaptive finite element mesh and mesh node coordinates are obtained. The measured evolution degree value Φ of the optimized monitoring point set is then read. measured Data quality checks were performed, and outliers deviating from the mean by 3σ were removed. Anisotropic Kriging interpolation was used with varying range parameters: longitudinal 50m (along the flow direction), lateral 20m, and vertical 10m. The interpolation uncertainty σ was calculated. kriging In sparse monitoring areas (nearest monitoring point > 30m), low confidence levels are marked; the initial evolution degree field Φ0(x, y, z) and confidence distribution map are obtained. The initial evolution degree field, evolution propagation equation coefficient field, and adaptive finite element mesh are read to construct a discretized equation set: [M]d 2 Φ / dt 2 +[C]dΦ / dt+[K]Φ=[F], where [M], [C], and [K] are the mass, damping, and stiffness matrices, respectively; [F] is the external load (source term) vector; implicit time integration is used (Newmark-β method, β=0.25, γ=0.5); convergence criterion ||Φ is set. (n+1) -Φ n ‖ / ‖Φ n ||<10 -4 , where Φ (n+1) Let be the evolution degree value vector after the (n+1)th iteration; the spatiotemporal evolution degree field Φ(x, y, z, t) is obtained by iterative solution. Read the measured values ​​of the spatiotemporal evolution degree field and the optimized monitoring point set, and calculate the reconstruction error e. i =|Φ reconstructed -Φ measured | / Φ measured , where Φ reconstructedThe evolution degree value is reconstructed; cross-validation is performed, and one monitoring point is removed and reconstructed sequentially, and the RMSE (root mean square of reconstruction error) is calculated; uncertainty propagation is analyzed, and a ±20% disturbance band is superimposed in the low confidence region; the overall evolution degree distribution Φ(x, y, z, t) and reconstruction uncertainty cloud map are output. The overall evolution degree distribution, the corrected concrete strength distribution map, and the mutation risk index are read, and the comprehensive health index H=w1×(1-Φ)+w2×(f) is calculated. c / f design )+w3×(1-P risk With weights w1=0.4, w2=0.35, and w3=0.25, a comprehensive health index distribution is obtained. This distribution is then divided into five levels: H≥0.85 for healthy, 0.70≤H<0.85 for basically healthy, 0.55≤H<0.70 for sub-healthy, 0.40≤H<0.55 for diseased, and H<0.40 for severely diseased. A color-coded overall health status distribution map is generated. Evolutionary state parameters, evolution rate distribution, and mutation risk indicators are read to establish a time series prediction model. The evolution degree Φ(t+Δt) for the next 30 / 90 / 180 days is extrapolated to identify time windows that may exceed the critical threshold, resulting in a risk evolution prediction curve and early warning time nodes. The system reads the overall health status distribution map, warning time points, and mutation type labels, and generates differentiated warnings based on risk levels: blue warning (within 180 days) for routine monitoring, yellow warning (within 90 days) for intensive monitoring, orange warning (within 30 days) for maintenance preparation, and red warning (within 7 days) for emergency response, outputting a tiered warning plan. It integrates the overall health status distribution map, risk evolution prediction curve, tiered warning plan, and all key parameters to generate a structural health assessment report including current status diagnosis, evolution trend analysis, risk level assessment, and response recommendations.

[0136] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A method for assessing the structural health of long-distance underground culverts, characterized in that, include: Based on acoustic and optical detection data, the system performs fusion processing to identify and output structural defect data that includes soil type, defect type, and geometric parameters. Based on pre-stored underwater rebound and core sampling test data, combined with structural defect data, underwater-to-dry conversion and correction of strength values ​​are performed to determine the corrected structural strength and material degradation parameters. By combining structural defect data, material degradation parameters, and pre-stored historical monitoring data, structural evolution modeling and mutation risk identification are performed to obtain the evolution characteristics and mutation risks of the structure. By comprehensively considering the corrected structural strength, evolutionary characteristics, and mutation risk, a full-field state assessment is conducted to generate structural health assessment and risk warning results. To obtain the evolutionary characteristics and mutation risks of the structure, including: By integrating structural defect data, material degradation parameters, and historical monitoring data, an evolution function coupled with multi-physics fields is constructed and solved to generate an evolution degree time series that quantifies the cumulative evolution degree of the structure. Differential analysis of the evolution time series identifies potential abrupt change moments that characterize risk transitions. Based on the evolution rate time series and the spatial location of structural defect data, an evolution rate distribution map is derived and generated; The evolutionary time series, potential mutation times, and evolution rate distribution map together constitute the evolutionary characteristics and mutation risks of the structure; Generate evolution time series, including: From historical monitoring data, extract physical field indicators including at least hydraulic gradient, seepage flow and soil strain, and calculate their respective time increment sequences; Based on soil type and material degradation parameters, the evolution weight coefficients corresponding to each physical field index are dynamically calibrated. The time increment sequences are weighted and fused using evolution weight coefficients to obtain an evolution increment sequence that reflects the instantaneous evolution rate of the structure. Numerical integration is performed on the evolution increment sequence along the time dimension to generate the evolution degree time series.

2. The method according to claim 1, characterized in that, Dynamically calibrated evolution weight coefficients include: Based on the soil type, assign basic weight coefficients to each physical field index; Structural degradation degree indicators are extracted from material degradation parameters, and at least one coefficient value in the basic weight coefficients is corrected accordingly to obtain the adjusted weights. The adjusted weights are normalized to generate evolution weight coefficients.

3. The method according to claim 1, characterized in that, Identify potential abrupt change moments that characterize risk transitions, including: Solve for the first and second derivative sequences of the evolution degree time series; A time series of mutation sensitivity is constructed by calculating the ratio of the absolute values ​​of the second derivative sequence and the first derivative sequence at each time step. Each sensitivity value in the mutation sensitivity time series is compared with a preset threshold, and the moments that exceed the threshold are identified as potential mutation moments.

4. The method according to claim 3, characterized in that, Also includes: For each identified potential mutation moment, acoustic data segments within a predetermined time window before and after that potential mutation moment are extracted from the acoustic detection data; Perform spectral analysis on the acoustic data segment to decode whether there are any acoustic anomaly modes associated with a pre-defined specific failure mechanism; If an acoustic anomaly pattern exists, the potential mutation moment is upgraded to a verified mutation risk, and the mutation risk is classified according to its cause based on the acoustic anomaly pattern.

5. The method according to claim 1, characterized in that, Determine the corrected structural strength and material degradation parameters, including: Based on core sampling test data, an underwater-dry land integrated conversion coefficient was constructed and calibrated. By applying the underwater-dry land integrated conversion coefficient, the underwater rebound is converted into a preliminary dry land equivalent strength value; Based on the defect type and geometric parameters, the preliminary dry equivalent strength value is locally corrected to obtain the corrected structural strength. Based on the comparison with the preset design strength, material degradation parameters are generated.

6. The method according to claim 5, characterized in that, Construct and calibrate the underwater-dry land integrated conversion coefficient, including: For the physical effects of the underwater environment, quantify and determine independent environmental correction factors, including at least two of the following: Pressure correction factor characterizing the effect of hydrostatic pressure on rebound value; Temperature correction factor characterizing the effect of water temperature on the elastic modulus of concrete; The scouring correction factor characterizes the effect of water flow scouring on the surface hardness of concrete; The media coupling coefficient characterizing the effect of water medium on the energy transfer of rebound impact; By coupling and integrating all environmental correction factors, an underwater-dry land integrated conversion coefficient is generated.

7. The method according to claim 5, characterized in that, To obtain the corrected structural strength, including: For each defect in the structural defect data, a three-dimensional influence field model is constructed based on its geometric parameters, showing that the influence intensity decreases with increasing spatial distance. Based on the type of defect, a basic strength reduction factor is set to characterize the degree of its maximum impact. By integrating the three-dimensional influence field model with the foundation strength reduction coefficient, a defect correction coefficient field that varies continuously in space is calculated. The defect correction coefficient field is applied to the initial dry equivalent strength value to obtain the corrected structural strength.

8. The method according to claim 1, characterized in that A full-field state assessment is performed, including a full-field reconstruction phase to generate a continuous state distribution, specifically: The evolutionary parameters contained in the evolutionary characteristics and mutation risks are set as sparse state monitoring points in the three-dimensional space of the culvert. Construct physical constraint equations to describe the spatiotemporal evolution of the propagation of the evolutionary state in the structural medium; Using data from sparse state monitoring points as the solution boundary or initial condition, the physical constraint equations are numerically solved to generate a full-field evolution distribution map that continuously covers the entire culvert.

Images