Bridge detection method and system based on digital twin technology

Abstract

The invention discloses a bridge detection method and system based on a digital twin technology, and relates to the field of bridge structure health monitoring. The method comprises the following steps: acquiring a strain distribution value, a vibration spectrum value and an environmental load spectrum value in real time through a sensor network deployed in a physical bridge, generating a structural response data set, and synchronizing the structural response data set to a digital twinborn body; calculating a damage index value and an accumulated damage quantity value based on the structural response data set; inputting the damage index value and the accumulated damage quantity value into a preset safety criterion, and calculating a safety margin coefficient value and a failure risk grade value; calculating a residual life prediction value based on the safety margin coefficient value and the environmental load spectrum value, and synchronously correcting a degradation rate value of the digital twin; and generating a priority maintenance instruction according to the failure risk grade value, the residual life prediction value and the safety margin coefficient value, and feeding back maintenance effect data to the digital twinborn body to complete updating after execution. The bridge operation and maintenance efficiency and safety are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge structure health monitoring, and in particular to a bridge detection method and system based on digital twin technology. Background Art

[0002] Current bridge structure health monitoring mainly relies on regular manual inspections and offline sensor data analysis, which has significant limitations.

[0003] Manual inspections are highly subjective and time-consuming, making it difficult to capture sudden damage; traditional vibration detection methods are easily affected by environmental noise and lack the accuracy to identify early damage such as microcracks; and visual inspection technology cannot quantify the degree of internal structural deterioration.

[0004] Existing digital twin applications primarily focus on the design phase, but suffer from three major flaws in operations and maintenance: first, multi-source data is processed in isolation, lacking dynamic correlation analysis; second, degradation models rely on static empirical formulas without incorporating real-time state correction parameters; and third, maintenance decisions are disconnected from the twin model, resulting in a broken "monitoring-assessment-maintenance" closed loop. This results in delayed warnings of bridge safety hazards, significant deviations in lifespan predictions, and irrational allocation of maintenance resources. There is an urgent need for an intelligent detection system that integrates real-time data-driven development, model self-optimization, and decision-making feedback. Summary of the Invention

[0005] The present invention proposes a bridge detection method based on digital twin technology, comprising: The sensor network deployed on the physical bridge collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time, generates a structural response dataset, and synchronizes it to the digital twin. Calculate the damage index value and cumulative damage value based on the structural response data set of the digital twin; Input the damage index value and the cumulative damage value into the preset safety criteria to calculate the safety margin coefficient value and the failure risk level value; Based on the safety margin coefficient value and environmental load spectrum value, the remaining life prediction value is calculated and the degradation rate value of the digital twin is simultaneously corrected; Priority maintenance instructions are generated based on the failure risk level value, remaining life prediction value and safety margin coefficient value. After execution, the maintenance effect data is fed back to the digital twin to complete the update.

[0006] The bridge inspection method based on digital twin technology described above includes the following sub-steps: a sensor network deployed on the physical bridge collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time, generates a structural response dataset, and synchronizes it to the digital twin. A high-density optical fiber strain sensor array is used to collect the three-dimensional strain distribution values ​​of the bridge main beam web, pier cap and cable tower anchorage area, and construct a stress field cloud map; A distributed accelerometer network is used to monitor the structural vibration response in real time, and the vibration spectrum characteristics such as fundamental frequency and damping ratio are extracted through fast Fourier transform; Integrate monitoring data from temperature and humidity sensors, anemometers, and dynamic weighing systems to generate multi-dimensional environmental load spectrum values ​​including temperature gradient, wind load spectrum, and traffic load time history; The edge computing gateway is used to perform spatiotemporal data registration to form a structural response dataset with spatiotemporal labels and synchronize it to the digital twin.

[0007] The bridge inspection method based on digital twin technology as described above, wherein the damage index value and the cumulative damage value are calculated based on the structural response data set of the digital twin, includes the following sub-steps: A local micro-strain field model is established based on the strain distribution value, and the fatigue damage index value of the key components is calculated by combining the material SN curve and Miner linear cumulative damage criterion; Based on the modal parameter offset of the vibration spectrum value, the cumulative damage value of the structural stiffness degradation is quantified through the frequency domain decomposition algorithm; By correlating historical inspection data with real-time responses, a time series prediction model is used to dynamically update the cumulative damage value and generate a full life cycle degradation curve.

[0008] The bridge inspection method based on digital twin technology described above, in which historical inspection data is correlated with real-time responses, a time series prediction model is used to dynamically update the cumulative damage value and generate a full life cycle degradation curve, includes the following sub-steps: Extract the strain distribution peak sequence and vibration fundamental frequency attenuation data stored in the digital twin over the years; By integrating real-time monitoring values ​​with historical databases through long-short-term memory networks, the cumulative damage value considering the material aging effect is reconstructed; The statistical significance of the cumulative damage value is verified based on the Bayesian update mechanism, and the exponential degradation parameters in the fatigue damage model are dynamically calibrated.

[0009] The bridge inspection method based on digital twin technology described above includes the following sub-steps: inputting the damage index value and the cumulative damage value into a preset safety criterion, and calculating the safety margin coefficient value and the failure risk level value. According to the spatial distribution of damage index values, the safety margin coefficient value of each area is calculated by matching the preset material strength threshold matrix; Combined with the spatiotemporal evolution characteristics of the cumulative damage value, a fuzzy comprehensive evaluation model is used to generate five levels of failure risk from low to high; The coverage completeness of the risk level determination logic is verified through Monte Carlo simulation to eliminate the risk of missed determination.

[0010] The bridge inspection method based on digital twin technology described above, wherein the remaining life prediction value is calculated based on the safety margin coefficient value and the environmental load spectrum value, and the degradation rate value of the digital twin is simultaneously corrected, includes the following sub-steps: The extreme working condition distribution of coupled environmental load spectrum values ​​and the safety margin coefficient attenuation curve are used to predict the remaining life based on the Paris crack growth law; Reversely optimize the time-varying degradation rate value of the digital twin based on the remaining life prediction results and establish an adaptive calibration mechanism for material performance degradation; Cross-validation between twins and physical entities ensures that the prediction accuracy of the degradation model is controlled within the ±5% error band.

[0011] The bridge inspection method based on digital twin technology described above includes the following sub-steps: generating priority maintenance instructions based on the failure risk level, the remaining life prediction value, and the safety margin coefficient value; and feeding back maintenance effect data to the digital twin to complete the update after execution. Based on the spatial heat map of failure risk level values, high-risk areas requiring emergency treatment and observation and monitoring areas are divided; Combined with the remaining life prediction critical value, an intelligent maintenance instruction set including construction timing and resource allocation is generated; The data on changes in structural dynamic characteristics after maintenance is collected through the IoT terminal, and closed-loop feedback is fed back to the digital twin to verify the effectiveness of the damage model correction.

[0012] The present invention also proposes a bridge detection system based on digital twin technology, comprising: Data acquisition and synchronization module: This module collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time through a sensor network deployed on the physical bridge, generates a structural response dataset, and synchronizes it to the digital twin. Structural health assessment module: Calculates damage index and cumulative damage values ​​based on the structural response dataset of the digital twin; Life prediction and state calibration module: Input the damage index value and cumulative damage value into the preset safety criteria to calculate the safety margin coefficient value and failure risk level value; Based on the safety margin coefficient value and environmental load spectrum value, calculate the remaining life prediction value and simultaneously correct the degradation rate value of the digital twin; Maintenance decision-making and closed-loop feedback module: Generates priority maintenance instructions based on the failure risk level value, remaining life prediction value and safety margin coefficient value, and feeds back the maintenance effect data to the digital twin after execution to complete the update.

[0013] The present invention also provides a computer storage medium, comprising: at least one memory and at least one processor; a memory for storing one or more program instructions; A processor is used to run one or more program instructions to execute any of the above-mentioned bridge detection methods based on digital twin technology.

[0014] The beneficial effects achieved by the present invention are as follows: A high-fidelity digital twin is constructed using real-time collected strain distribution values, vibration spectrum values, and environmental load spectrum values, significantly improving the accuracy of structural status perception. Damage index values ​​and cumulative damage values ​​are collaboratively calculated based on multi-source data to accurately identify hidden damage such as microcrack extension and stiffness degradation. Safety margin coefficient values ​​and failure risk level values ​​are dynamically generated in combination with preset safety criteria, realizing the transition from passive detection to active warning. The remaining life is predicted through the coupling analysis of environmental load spectrum values ​​and safety margins, and the digital twin degradation model is corrected in real time, greatly improving the reliability of life assessment. Priority maintenance instructions are generated based on risk level, life prediction, and safety margin to optimize resource allocation and verify maintenance effects, forming a "monitoring-assessment-decision-making-verification" closed loop, effectively extending the service life of bridges, reducing the risk of sudden accidents, and providing core support for smart infrastructure management and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0016] Figure 1 This is a flow chart of a bridge detection method based on digital twin technology provided in an embodiment of the present application.

[0017] Figure 2 This is a schematic diagram of a bridge detection system based on digital twin technology provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following is a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0019] Example 1

[0020] like Figure 1As shown, the embodiment of the present application provides a bridge detection method based on digital twin technology, including: Step S1: The sensor network deployed on the physical bridge collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time, generates a structural response data set, and synchronizes it to the digital twin; Specifically, a high-density fiber optic sensor array is used to acquire three-dimensional deformation data of key bridge parts in real time, generating a dynamic stress field cloud map. A network of deployed accelerometers captures structural vibration signals, and spectrum analysis is used to analyze the fundamental frequency and damping characteristics. Time series data on temperature gradients, wind loads, and traffic loads acquired by environmental monitoring equipment are simultaneously integrated. Edge computing nodes are used to achieve spatiotemporal calibration of multi-source information, construct a structural response set with unified spatiotemporal labels, and update the digital twin in real time. This includes the following sub-steps: Step S11: collecting three-dimensional strain distribution values ​​of the bridge main beam web, pier cap and cable tower anchorage area through a high-density optical fiber strain sensor array to construct a stress field cloud map; A high-density array of fiber-optic strain sensors is deployed in key stress-bearing areas of the bridge. This array continuously collects three-dimensional strain components on the surface or within the structure at a preset spatial resolution. The raw optical signals collected are converted into digital strain data by a demodulator. A spatial interpolation algorithm is then used to generate a detailed stress field distribution cloud map covering the monitored area, visually reflecting the stress levels and gradients of the structure under load.

[0021] Step S12: using a distributed accelerometer network to monitor the structural vibration response in real time, and extracting vibration spectrum characteristic values ​​such as fundamental frequency and damping ratio through fast Fourier transform; A distributed accelerometer network is deployed at sensitive locations along the bridge structure's key modes. The accelerometers simultaneously record the structure's vibration acceleration time history data at a high sampling rate under environmental excitation or traffic loads. After preprocessing the raw acceleration signals, they are converted to the frequency domain using a fast Fourier transform algorithm. This accurately identifies key vibration spectrum characteristics, including the structure's primary fundamental frequency, higher-order harmonic frequencies, and modal damping ratios calculated using the half-power bandwidth method or random subspace identification. This is used to assess the structure's overall dynamic characteristics and health.

[0022] Step S13: Integrate the monitoring data of the temperature and humidity sensor, the anemometer and the dynamic weighing system to generate a multi-dimensional environmental load spectrum value including temperature gradient, wind load spectrum and traffic load time history; A variety of environmental and load sensors are integrated at key locations on and around the bridge. A network of temperature and humidity sensors is deployed across different sections of the bridge structure to measure the temperature gradient distribution and ambient humidity within the structure. The temperature gradient field is calculated using the following formula:

[0023] in, Indicates the effect of the mass properties of the substance itself on temperature changes during heat transfer; represents the specific heat capacity at constant pressure; represents the partial derivative of temperature with respect to time; represents the heat conduction term, represents the gradient operator; represents the temperature gradient; It represents thermal conductivity, which reflects the ability of a substance to conduct heat; represents the solar radiation flux; represents the convective heat transfer coefficient; represents the temperature of the research object; represents the far-field temperature of the environment; It represents emissivity, which describes the ability of an object's surface to emit radiation energy. Its value ranges from 0 to 1, reflecting the degree to which the object is close to blackbody radiation. It is used to calculate the heat transfer of thermal radiation and is a key constant in the blackbody radiation law; represents the effective radiation temperature of the sky; Anemometers are installed at unobstructed locations above the bridge deck and on the tops of cable towers to capture the wind speed, direction, and spectrum characteristics acting on the structure in real time. The wind load spectrum is expressed using the following formula:

[0024] in, represents the power spectrum density of fluctuating wind; Indicates frequency; It represents an empirical constant, which is related to factors such as site and topography and is used to fit the actual wind spectrum characteristics; represents the friction speed; represents the characteristic frequency; represents the coefficient related to the wind profile characteristics; Indicates height The average wind speed at Indicates the height of the calculated wind load position relative to the ground; represents the experience correction coefficient; represents the phase perturbation function.

[0025] The dynamic weighing system is embedded in a specific lane of the bridge deck pavement layer to record the axle weight, wheelbase, speed and traffic density of passing vehicles, generating accurate traffic load time history data. The following formula is used to represent the random process of traffic load:

[0026] in, Represents the traffic load random process at time Response; Indicates the deadline Total number of arriving load events; Indicates the sequence number of the load event; Indicates the The amplitude of each load event; Indicates the The arrival time of each load event; Indicates: Velocity parameters associated with each load event; Indicates the Other characteristic parameters of each load event; Indicates the The time history function of each load event.

[0027] After synchronous collection, the above monitoring data are integrated to form a multi-dimensional environmental load spectrum value dataset including spatial temperature field, wind load time-frequency characteristics and traffic load time history.

[0028] Step S14: Use the edge computing gateway to perform spatiotemporal data registration to form a structural response dataset with spatiotemporal labels and synchronize it to the digital twin; The gateway receives raw data streams from the aforementioned sensors and utilizes a high-precision timing module to achieve strict time synchronization across all data channels, ensuring temporal consistency. Simultaneously, each data point is assigned a corresponding spatial location label based on the precise spatial coordinates of the sensors on the physical bridge. The gateway's built-in preprocessing module performs preliminary verification of the raw data, standardizes its format, and removes invalid data. After completing spatiotemporal registration, the multi-source heterogeneous data is packaged into a structural response dataset with spatiotemporal labels. This dataset is transmitted in real-time or near-real-time via a high-speed communication link and is synchronously updated to the digital twin platform located on a remote server.

[0029] Step S2: Calculate the damage index value and the cumulative damage value based on the structural response data set of the digital twin; Specifically, the real-time damage index of core components is calculated based on the local deformation field model combined with the material fatigue characteristics; the structural stiffness degradation is analyzed using the vibration frequency characteristic offset, and the cumulative damage value is quantified using a frequency domain algorithm; historical detection data is simultaneously integrated, and the dynamic damage accumulation under the material aging effect is reconstructed based on the time series prediction model; finally, the statistical reliability of the damage amount is verified using a Bayesian mechanism, the fatigue degradation parameters are optimized in real time, and the full life cycle performance evolution curve is generated. Specifically, the following sub-steps are included: Step S21: Establish a local microstrain field model based on the strain distribution value, and calculate the fatigue damage index value of the key component by combining the material SN curve and Miner linear cumulative damage criterion; High-density strain sensor measurements of designated critical components are extracted from the structural response dataset updated in real time by the digital twin. A local microstrain field distribution map reflecting the critical path on the component surface or internally is constructed based on a spatial interpolation algorithm. Based on the standardized SN fatigue characteristic curve of the component material, the number of fatigue life cycles corresponding to different stress amplitudes is determined. Applying the Miner linear cumulative damage criterion, the stress time history data obtained through real-time monitoring and processed by the rainflow counting method is decomposed into stress cycle blocks of different amplitudes. The damage ratio caused by each stress cycle block is calculated using the formula: Multiaxial stress correction formula:

[0030] in, The average stress and stress gradient are corrected. The equivalent stress amplitude of each stress state; represents the stress amplitude; represents the mean stress; Indicates ultimate tensile strength; represents the multiaxial correction index; represents the gradient sensitivity coefficient; Represents the Frobenius norm of the strain gradient tensor, which is used to measure the magnitude of the strain gradient; Nonlinear SN curve equation:

[0031] in, Indicates the Fatigue life under a stress state; represents the Basquin parameter, represents the fatigue strength coefficient; represents the fatigue strength index; Indicates the equivalent stress amplitude; represents the fatigue limit; represents the curvature correction parameter; represents the Laplacian norm of the strain gradient tensor; Indicates the reference strain value.

[0032] Damage ratio calculation formula:

[0033] in, Indicates the The damage ratio corresponding to each stress cycle block; Indicates the number of cycles of this stress cycle block; Indicates the fatigue life under the corresponding stress state; represents the strain gradient tensor The Frobenius norm of Represents the strain gradient tensor, the elements contain strain components Coordinates ; represents the damage activation function; Indicates the equivalent stress amplitude; represents the fatigue limit; The real-time fatigue damage index value of the key component in the current monitoring period is obtained by accumulating block by block.

[0034] Step S22: quantifying the cumulative damage value of the structural stiffness degradation by a frequency domain decomposition algorithm based on the modal parameter offset of the vibration spectrum value; A high-precision frequency-domain decomposition algorithm is used to identify modal parameters of vibration signals, obtaining the primary modal frequencies, damping ratios, and mode shapes of the structure in its current state. The identified current modal frequencies are compared with the corresponding modal frequencies stored in the digital twin benchmark model, and their relative offsets are calculated. Combined with structural mechanics analysis, a quantitative mapping relationship is established between modal frequency offsets and overall or local structural stiffness degradation. The identified multi-order modal frequency offsets are then converted into cumulative damage values ​​reflecting stiffness degradation of the structure as a whole or in key subsystems.

[0035] Step S23: Correlate historical detection data with real-time responses, dynamically update the cumulative damage value using a time series prediction model, and generate a full life cycle degradation curve; Step S231: extract the strain distribution peak sequence and vibration fundamental frequency attenuation data stored in the digital twin over the years; From the digital twin's integrated long-term historical database, a time series of typical strain distribution peak data for designated key components recorded during important inspections or specific events is extracted. Simultaneously, a series of first-order or major-order fundamental vibration frequency measurements of the structure's entire structure are extracted at corresponding time points, forming an attenuation data chain reflecting the fundamental frequency's temporal changes.

[0036] Step S232: Reconstruct the cumulative damage value taking into account the material aging effect by fusing the real-time monitoring value with the historical database through the long short-term memory network; The obtained real-time damage index values ​​for key components and the current cumulative damage values ​​for the structure are spatially and temporally aligned and integrated with corresponding historical data sequences. A time series analysis algorithm capable of capturing long-term dependencies is applied to establish a damage evolution analysis mechanism that integrates historical trends with real-time status information. Through algorithmic processing, cumulative damage predictions for key components and the overall structure, under the background of material aging, are dynamically reconstructed and updated to better align with actual long-term evolutionary laws.

[0037] Step S233: verifying the statistical significance of the cumulative damage value based on the Bayesian update mechanism, and dynamically calibrating the exponential degradation parameters in the fatigue damage model; Using the Bayesian statistical inference framework, the predicted cumulative damage value of the aging effect is compared with the prediction results of the physics-based fatigue damage model and stiffness degradation model. Taking the aforementioned fused prediction value as new observational evidence, the Bayesian update mechanism is used to calculate the posterior probability distribution of the model prediction to evaluate the statistical significance and uncertainty level of the current model prediction results. Based on the posterior analysis results, the key time-varying parameters in the fatigue damage model and stiffness degradation model are dynamically calibrated and optimized. The calibrated parameters are fed back and updated in real time to the core computational model of the digital twin to ensure that the model continues to reflect the true degradation state of the structure. Finally, based on the continuously updated model and fused data, a key performance evolution curve reflecting the entire life cycle of the bridge structure from its construction to the present and predicted to the future is generated and output.

[0038] Step S3: Input the damage index value and the cumulative damage value into the preset safety criteria to calculate the safety margin coefficient value and the failure risk level value; Specifically, the cutting resistance deviation is calculated by real-time analysis of the acoustic signal spectrum characteristics and strain mutation points, and the kinetic energy compensation value is derived according to the preset mapping relationship, thereby dynamically adjusting the driving torque and feed rate. The specific sub-steps include: Step S31: Calculate the safety margin coefficient value of each region according to the spatial distribution of the damage index value and matching the preset material strength threshold matrix; The system spatially aligns and compares the damage index values ​​generated by the sensor network data with the predefined material strength threshold matrix in the three-dimensional space of the bridge structure. A spatial interpolation algorithm is used to map the discrete damage index to a continuous threshold matrix grid. For each assessment unit, the system uses the ratio of the damage index value to the material strength threshold at the corresponding location, combined with a weighted algorithm that considers the effects of local stress concentration and damage coupling, to calculate the safety margin coefficient value that reflects the residual bearing capacity of the structure in that area. The final calculation of the safety margin coefficient is expressed using the following formula:

[0039] in, Indicates the safety margin factor to be calculated in the end; Express about Fatigue limit; represents the damage correlation function; represents a numerical stability term; Represents Some kind of correction and control function related to it; Indicates the minimum stress reference value; represents a scale parameter used to define the range of stress difference; It represents an exponential function term, which is used to consider the effect of damage gradient on safety margin.

[0040] Step S32: combining the spatiotemporal evolution characteristics of the cumulative damage value, using a fuzzy comprehensive evaluation model to generate five levels of failure risk level values ​​from low to high; Based on these spatiotemporal evolutionary characteristics, the system applies a fuzzy logic assessment framework, incorporating key indicators such as the cumulative damage magnitude, its rate of change, and the breadth of its spatial distribution as input variables. Through a fuzzy inference process, the system comprehensively considers the contribution of each input variable to the overall failure probability. The system then defuzzifies the fuzzy output and ultimately maps it into five discrete levels, ranging from "very low risk" to "very high risk." This intuitively quantifies the potential failure probability of key bridge components or the entire structure under expected service conditions.

[0041] Step S33: Verify the coverage completeness of the risk level determination logic through Monte Carlo simulation to eliminate the risk of missed determination; A parameterized probabilistic model was constructed to account for the randomness of material properties, uncertainty in load effects, errors in test data, and variability in boundary conditions. Based on this model, large-scale random sampling simulations were conducted within the risk level determination logic framework to simulate the bridge's response under various possible operating conditions and output corresponding risk levels. By statistically analyzing the massive simulation results, the system focused on examining whether high-risk scenarios, such as critical damage states and rare load events, could be accurately identified and classified into corresponding high-risk levels. In particular, the determination logic's ability to discern potential "gray areas" or complex coupled damage patterns was verified.

[0042] Step S4: Calculate the remaining life prediction value based on the safety margin coefficient value and the environmental load spectrum value, and simultaneously correct the degradation rate value of the digital twin; Specifically, the remaining life prediction value is calculated based on the safety margin coefficient value and the environmental load spectrum value, and the degradation rate value of the digital twin is simultaneously corrected. Specifically, the extreme working conditions of the environmental load spectrum are coupled with the safety margin attenuation curve, and the remaining life is calculated based on the Paris law. The twin material degradation rate parameters are reversely optimized based on the deviation of the prediction value. The twin prediction is cross-validated with the physical monitoring data to control the prediction accuracy within ±5%. If it exceeds the limit, calibration is triggered. The specific sub-steps include the following: Step S41: coupling the extreme working condition distribution of the environmental load spectrum value and the safety margin coefficient attenuation curve, and predicting the remaining life prediction value based on the Paris crack growth law; Identify the load distribution characteristics that characterize extreme service conditions in the environmental load spectrum and couple them with the attenuation curve of the safety margin coefficient of key structural parts obtained through real-time monitoring over service time. Based on the structural damage evolution law described by the Paris crack growth theory, the current crack size, material fracture toughness, and the coupled load-resistance interaction characteristics are combined to construct a remaining life prediction equation, which is expressed as follows:

[0043] in, Indicates the remaining lifespan; represents the initial crack length; represents the critical crack length; represents a small increment of crack growth; represents the material constant; represents the effective stress intensity factor range; Represents the geometric factor, which is used to modify the stress intensity factor; Indicates the crack growth rate sensitivity index of the material; Indicates coefficients related to materials and environment; Indicates the rate at which the sliding displacement rate and other parameters change with the crack length during crack propagation; A time representation indicating the remaining life; represents the equivalent cycle frequency; It represents the time period for statistics or calculation of equivalent cycle frequency, and is a time interval used for integral calculation; Represents a quantity related to the load history; Indicates its rate of change over time.

[0044] By solving this equation, the predicted remaining service life of key structural parts under the expected service environment is calculated.

[0045] Step S42: reversely optimize the time-varying degradation rate value of the digital twin based on the remaining life prediction result, and establish an adaptive calibration mechanism for material performance degradation; The calculated predicted remaining life value is compared with the remaining life value predicted by the current simulation of the digital twin. Based on this predicted deviation, the possible error in the time-varying rate parameters of the simulated material performance degradation in the digital twin is reversed. This deviation information is used to dynamically generate correction factors for the material performance degradation rate parameters within the twin. By applying the correction factors to the twin's degradation rate parameters, online, adaptive calibration of the digital twin's material performance degradation process is achieved, ensuring that the degradation trajectory of the twin simulation is consistent with the degradation trend revealed by actual monitoring.

[0046] Step S43: Ensure that the prediction accuracy of the degradation model is controlled within the ±5% error band through cross-validation between the twin and the physical entity; Establish a normalized cross-validation mechanism between the digital twin's predicted data and the measured data of the physical bridge entity. Regularly perform quantitative comparisons between the key responses or derived indicators of the twin based on the current degradation model and load input simulation calculations and the actual sensor monitoring data at the corresponding position and time. Use statistical process control methods to continuously monitor the relative error between the predicted value and the measured value. Once the error exceeds the preset ±5% allowable error band, the degradation rate reverse optimization calibration process in step S42 is automatically triggered to iteratively correct the twin parameters. Through this closed-loop feedback mechanism, the prediction accuracy of the digital twin's simulation of the degradation process of bridge structure material performance is continuously improved and maintained within an acceptable range for engineering.

[0047] Step S5: Generate priority maintenance instructions based on the failure risk level value, remaining life prediction value, and safety margin coefficient value. After execution, the maintenance effect data is fed back to the digital twin to complete the update; Specifically, high-risk and monitoring areas are divided based on the failure risk heat map; maintenance instruction sets are generated based on the risk level, remaining life, and safety margin; and post-maintenance power parameters are collected through the Internet of Things and fed back to the digital twin to verify the effect and update the status. The specific sub-steps include: Step S51: Based on the spatial heat map of the failure risk level value, divide the high-risk area requiring emergency treatment and the observation and monitoring area; Based on the digital twin-generated heat map of the spatial distribution of bridge structure failure risk levels, continuous color blocks representing extremely high risk levels and their spatial boundaries are identified. By analyzing the color gradient of the heat map, a risk level threshold is set, and areas exceeding this threshold are automatically designated as high-risk areas. These areas are then marked as priority areas requiring immediate engineering intervention, such as reinforcement or repair. Furthermore, areas with moderate risk levels and relatively discrete spatial distribution are identified and designated as observation and monitoring areas. These areas are marked as requiring increased regular inspections and monitoring frequency but not requiring immediate engineering intervention. The resulting demarcation is then overlaid onto the bridge digital twin model as a visual layer.

[0048] Step S52: Generate an intelligent maintenance instruction set including construction sequence and resource allocation in combination with the remaining life prediction critical value; For the identified high-risk areas, the corresponding predicted remaining life values ​​are extracted. A critical safety threshold for remaining life is set, and components or parts in high-risk areas with predicted remaining life values ​​below this threshold are identified as emergency response items. A comprehensive ranking is performed based on failure risk level, predicted remaining life values, and safety margin coefficient values, automatically generating a maintenance priority sequence. Based on this sequence, an optimization algorithm is used to generate a detailed construction schedule and resource allocation plan, taking into account factors such as the availability of engineering resources, construction process requirements, and traffic impacts. Ultimately, a structured, executable set of intelligent maintenance instructions is output to guide on-site maintenance operations.

[0049] Step S53: Collect data on changes in structural dynamic characteristics after maintenance through the IoT terminal, and feed it back to the digital twin in a closed loop to verify the effectiveness of the damage model correction; After the maintenance instruction set is executed, the IoT sensor network deployed at key locations on the bridge collects real-time data on the structural dynamic response of the bridge under operational loads. Key parameters of the post-maintenance structure, such as natural frequency, modal vibration shape, and damping ratio, are collected and calculated. The collected post-maintenance dynamic characteristic data set is transmitted to the digital twin platform via a secure communication link. The received measured data is compared and analyzed with the historical baseline data stored in the twin before maintenance. If the measured dynamic characteristic parameters meet or exceed the theoretical improvement model expected based on the maintenance measures, the maintenance measures are deemed effective, and this closed-loop verification of the correction logic and prediction accuracy of the damage identification and evolution model built into the digital twin after this maintenance intervention is conducted.

[0050] Example 2

[0051] like Figure 2 As shown, the second embodiment of the present application provides a bridge detection system based on digital twin technology, including: Data Acquisition and Synchronization Module 21: This module collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time through a sensor network deployed on the physical bridge, generates a structural response dataset, and synchronizes it to the digital twin. It includes the following submodules: Strain field acquisition submodule 211: collects three-dimensional strain distribution values ​​of the bridge main beam web, pier cap and cable tower anchorage area through a high-density optical fiber strain sensor array to construct a stress field cloud map; Vibration spectrum monitoring submodule 212: uses a distributed accelerometer network to monitor the structural vibration response in real time, and extracts vibration spectrum characteristic values ​​such as fundamental frequency and damping ratio through fast Fourier transform; Environmental load integration submodule 213: Integrates monitoring data from temperature and humidity sensors, anemometers, and dynamic weighing systems to generate a multidimensional environmental load spectrum value including temperature gradient, wind load spectrum, and traffic load time history; Edge synchronization gateway submodule 214: uses the edge computing gateway to perform spatiotemporal data registration, forms a structural response dataset with spatiotemporal labels and synchronizes it to the digital twin; Structural Health Assessment Module 22: Calculates damage index values ​​and cumulative damage values ​​based on the structural response dataset of the digital twin. It includes the following submodules: Fatigue damage index calculation submodule 221: establishes a local microstrain field model based on the strain distribution value, and calculates the fatigue damage index value of the key component by combining the material SN curve and Miner linear cumulative damage criterion; Stiffness degradation identification submodule 222: quantifying the cumulative damage value of structural stiffness degradation through frequency domain decomposition algorithm based on the modal parameter offset of the vibration spectrum value; Full life cycle prediction submodule 223: associates historical detection data with real-time responses, uses a time series prediction model to dynamically update the cumulative damage value and generate a full life cycle degradation curve; Life prediction and state calibration module 23: Inputs the damage index value and cumulative damage value into the preset safety criteria, calculates the safety margin coefficient value and failure risk level value; calculates the remaining life prediction value based on the safety margin coefficient value and environmental load spectrum value, and simultaneously corrects the degradation rate value of the digital twin; includes the following submodules: Safety margin field calculation submodule 231: calculates the safety margin coefficient value of each region according to the spatial distribution of the damage index value and the preset material strength threshold matrix; Risk dynamic assessment submodule 232: combining the spatiotemporal evolution characteristics of the cumulative damage value, a fuzzy comprehensive assessment model is used to generate five levels of failure risk level values ​​from low to high; Risk coverage verification submodule 233: Verify the coverage completeness of the risk level determination logic through Monte Carlo simulation to eliminate the risk of missed determination; Crack growth life prediction submodule 234: coupling the extreme working condition distribution of the environmental load spectrum value and the safety margin coefficient attenuation curve, and predicting the remaining life prediction value based on the Paris crack growth law; Adaptive calibration submodule 235: reversely optimizes the time-varying degradation rate value of the digital twin based on the remaining life prediction results, and establishes an adaptive calibration mechanism for material performance degradation; Closed-loop control submodule 236 ensures that the degradation model prediction accuracy is controlled within the ±5% error band through cross-validation between the twin and the physical entity; Maintenance decision and closed-loop feedback module 24: Generates priority maintenance instructions based on the failure risk level value, remaining life prediction value, and safety margin coefficient value. After execution, the maintenance effect data is fed back to the digital twin to complete the update. It includes the following submodules: Hierarchical treatment submodule 241: based on the spatial heat map of the failure risk level value, divide the high-risk area requiring emergency treatment and the observation and monitoring area; Decision optimization submodule 242: generates an intelligent maintenance instruction set including construction timing and resource allocation in combination with the remaining life prediction critical value; Twin effectiveness feedback submodule 243: collects data on changes in structural dynamic characteristics after maintenance through the IoT terminal, and feeds closed-loop feedback to the digital twin to verify the effectiveness of damage model modification; Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, comprising: at least one memory and at least one processor; The memory is used to store one or more program instructions; A processor is used to run one or more program instructions to execute a bridge detection method based on digital twin technology.

[0052] Corresponding to the above embodiment, an embodiment of the present invention provides a computer-readable storage medium, which contains one or more program instructions, and the one or more program instructions are used by a processor to execute a bridge detection method based on digital twin technology.

[0053] The embodiments disclosed in the present invention provide a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed on a computer, the computer executes the above-mentioned bridge detection method based on digital twin technology.

[0054] In the embodiments of the present invention, the processor may be an integrated circuit chip having signal processing capabilities. The processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0055] The methods, steps, and logic diagrams disclosed in the embodiments of the present invention can be implemented or executed. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of the present invention can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules within the decoding processor. The software modules can be located in a storage medium such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers, which are well-known in the art. The processor reads the information from the storage medium and, in conjunction with its hardware, completes the steps of the aforementioned method.

[0056] The storage medium may be a memory and may be, for example, a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memory.

[0057] Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.

[0058] Volatile memory may be random access memory (RAM), which is used as an external cache memory. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct RAM bus random access memory (DRRAM).

[0059] The storage media described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memory.

[0060] Those skilled in the art will appreciate that in one or more of the above examples, the functions described herein can be implemented using a combination of hardware and software. When software is used, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media includes any medium that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0061] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, and improvements made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A bridge detection method based on digital twin technology, characterized in that: include: The sensor network deployed on the physical bridge collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time, generates a structural response dataset, and synchronizes it to the digital twin. Calculate the damage index value and cumulative damage value based on the structural response data set of the digital twin; Input the damage index value and the cumulative damage value into the preset safety criteria to calculate the safety margin coefficient value and the failure risk level value; Based on the safety margin coefficient value and environmental load spectrum value, the remaining life prediction value is calculated and the degradation rate value of the digital twin is simultaneously corrected; Priority maintenance instructions are generated based on the failure risk level value, remaining life prediction value and safety margin coefficient value. After execution, the maintenance effect data is fed back to the digital twin to complete the update.

2. The bridge detection method based on digital twin technology according to claim 1 is characterized in that: The sensor network deployed on the physical bridge collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time, generates a structural response dataset, and synchronizes it to the digital twin. This involves the following sub-steps: A high-density optical fiber strain sensor array is used to collect the three-dimensional strain distribution values ​​of the bridge main beam web, pier cap and cable tower anchorage area, and construct a stress field cloud map; A distributed accelerometer network is used to monitor the structural vibration response in real time, and the vibration spectrum characteristics such as fundamental frequency and damping ratio are extracted through fast Fourier transform; Integrate monitoring data from temperature and humidity sensors, anemometers, and dynamic weighing systems to generate multi-dimensional environmental load spectrum values ​​including temperature gradient, wind load spectrum, and traffic load time history; The edge computing gateway is used to perform spatiotemporal data registration to form a structural response dataset with spatiotemporal labels and synchronize it to the digital twin.

3. The bridge detection method based on digital twin technology according to claim 1 is characterized in that: Based on the structural response dataset of the digital twin, the damage index value and the cumulative damage value are calculated, which includes the following sub-steps: A local micro-strain field model is established based on the strain distribution value, and the fatigue damage index value of the key components is calculated by combining the material SN curve and Miner linear cumulative damage criterion; Based on the modal parameter offset of the vibration spectrum value, the cumulative damage value of the structural stiffness degradation is quantified through the frequency domain decomposition algorithm; By correlating historical inspection data with real-time responses, a time series prediction model is used to dynamically update the cumulative damage value and generate a full life cycle degradation curve.

4. The bridge detection method based on digital twin technology according to claim 3 is characterized in that: Correlating historical inspection data with real-time responses, using a time series prediction model to dynamically update the cumulative damage value and generate a full life cycle degradation curve includes the following sub-steps: Extract the strain distribution peak sequence and vibration fundamental frequency attenuation data stored in the digital twin over the years; By integrating real-time monitoring values ​​with historical databases through long-short-term memory networks, the cumulative damage value considering the material aging effect is reconstructed; The statistical significance of the cumulative damage value is verified based on the Bayesian update mechanism, and the exponential degradation parameters in the fatigue damage model are dynamically calibrated.

5. The bridge detection method based on digital twin technology according to claim 1 is characterized in that: Input the damage index value and the cumulative damage value into the preset safety criteria to calculate the safety margin coefficient value and the failure risk level value, including the following sub-steps: According to the spatial distribution of damage index values, the safety margin coefficient value of each area is calculated by matching the preset material strength threshold matrix; Combined with the spatiotemporal evolution characteristics of the cumulative damage value, a fuzzy comprehensive evaluation model is used to generate five levels of failure risk from low to high; The coverage completeness of the risk level determination logic is verified through Monte Carlo simulation to eliminate the risk of missed determination.

6. The bridge detection method based on digital twin technology according to claim 1 is characterized in that: Based on the safety margin coefficient value and the environmental load spectrum value, the remaining life prediction value is calculated and the degradation rate value of the digital twin is simultaneously corrected. The following sub-steps are included: The extreme working condition distribution of coupled environmental load spectrum values ​​and the safety margin coefficient attenuation curve are used to predict the remaining life based on the Paris crack growth law; Inversely optimize the time-varying degradation rate value of the digital twin based on the remaining life prediction results and establish an adaptive calibration mechanism for material performance degradation; Cross-validation between twins and physical entities ensures that the prediction accuracy of the degradation model is controlled within the ±5% error band.

7. The bridge detection method based on digital twin technology according to claim 1 is characterized in that: Generate priority maintenance instructions based on the failure risk level, remaining life prediction, and safety margin coefficient. After execution, the maintenance effect data is fed back to the digital twin to complete the update. This includes the following sub-steps: Based on the spatial heat map of failure risk level values, high-risk areas requiring emergency treatment and observation and monitoring areas are divided; Combined with the remaining life prediction critical value, an intelligent maintenance instruction set including construction timing and resource allocation is generated; The data on changes in structural dynamic characteristics after maintenance is collected through the IoT terminal, and closed-loop feedback is fed back to the digital twin to verify the effectiveness of the damage model correction.

8. A bridge detection system based on digital twin technology, characterized in that: include: Data acquisition and synchronization module: This module collects strain distribution values, vibration spectrum values, and environmental load spectrum values ​​in real time through a sensor network deployed on the physical bridge, generates a structural response dataset, and synchronizes it to the digital twin. Structural health assessment module: Calculates damage index and cumulative damage values ​​based on the structural response dataset of the digital twin; Life prediction and state calibration module: inputs the damage index value and cumulative damage value into the preset safety criteria, calculates the safety margin coefficient value and failure risk level value; Based on the safety margin coefficient value and environmental load spectrum value, the remaining life prediction value is calculated and the degradation rate value of the digital twin is simultaneously corrected; Maintenance decision-making and closed-loop feedback module: Generates priority maintenance instructions based on the failure risk level value, remaining life prediction value and safety margin coefficient value, and feeds back the maintenance effect data to the digital twin after execution to complete the update.

9. A computer storage medium, characterized in that include: at least one memory and at least one processor; a memory for storing one or more program instructions; A processor is used to run one or more program instructions to execute a bridge detection method based on digital twin technology as described in any one of claims 1 to 7.

Images