High-altitude large steel structure corridor safety risk monitoring method and system and medium

Abstract

The invention relates to a high-altitude large steel structure corridor safety risk monitoring method and system and a medium, and belongs to the technical field of civil engineering structure health monitoring, and the monitoring method comprises the steps: collecting original monitoring data through a multi-source sensor group disposed at each node of a steel structure corridor; performing space-time alignment processing on the original monitoring data to generate a sensor data matrix; performing dynamic noise suppression on the sensor data matrix based on an environmental noise transfer function model, and outputting a pure signal matrix and a damage characteristic frequency band identifier; calculating a thermal stress sensitive factor, extracting a multi-physical field coupling feature associated with the damage feature frequency band identifier in the pure signal matrix, fusing to generate a high-dimensional damage feature tensor, inputting the high-dimensional damage feature tensor into a pre-constructed digital twinborn risk assessment model, and outputting an assessment result including a risk level label and a damage position coordinate; and matching the regulation and control strategy according to the risk level label, and generating a corresponding equipment control instruction. According to the invention, the scientificity and accuracy of risk decision can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of civil engineering structure health monitoring, and in particular to a method, system and medium for monitoring the safety risks of a large-scale high-altitude steel structure corridor. Background Art

[0002] With the acceleration of urbanization and the continuous emergence of high-rise buildings, large-span, high-altitude cantilevered steel corridors are widely used in commercial complexes, transportation hubs, and public buildings due to their excellent space utilization efficiency and aesthetics. However, these structures are exposed to complex external environments (such as wind loads, temperature fluctuations, and vibration disturbances) for a long time, making them susceptible to various potential threats such as fatigue damage, loose connections, and local buckling. If not detected and addressed in a timely manner, these threats can lead to serious safety accidents. Therefore, how to achieve comprehensive perception, accurate identification, and efficient response to these structures during their service life has become a research hotspot and technical difficulty in the field of civil engineering structural safety monitoring.

[0003] Currently, common structural health monitoring systems typically rely on sensors (such as strain gauges or accelerometers) for local state perception, supplemented by manual inspections or regular testing to determine structural health. However, when dealing with high-altitude, large-scale, multi-degree-of-freedom steel corridors, the actual monitoring signals from sensors are susceptible to environmental noise contamination. Furthermore, most systems lack the ability to deeply mine features after data collection, making it difficult to convert raw signals into high-dimensional damage indicators that can be used for structural performance assessment. This, in turn, impacts the scientific nature and accuracy of subsequent risk decision-making. Summary of the Invention

[0004] In order to improve the scientificity and accuracy of risk decision-making, this application provides a method, system and medium for monitoring the safety risks of large-scale high-altitude steel structure corridors.

[0005] In the first aspect, the present application provides a method for monitoring the safety risks of a large-scale high-altitude steel structure corridor, which adopts the following technical solutions: A method for monitoring safety risks of a high-altitude large-scale steel structure corridor, the monitoring method comprising: The original monitoring data is collected by a multi-source sensor group deployed at each node of the steel structure corridor; the original monitoring data includes strain signals, vibration signals, acoustic emission signals and temperature signals; Performing spatiotemporal alignment processing on the raw monitoring data to generate a time-synchronized sensor data matrix; Dynamically suppress the sensor data matrix based on a pre-trained environmental noise transfer function model, and output a clean signal matrix after noise reduction and a damage characteristic frequency band identifier; Calculating a thermal stress sensitivity factor based on the strain signal and the temperature signal, extracting multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fusing them to generate a high-dimensional damage characteristic tensor; The high-dimensional damage characteristic tensor is input into a pre-built digital twin risk assessment model. The digital twin risk assessment model locally refines the finite element mesh and updates the material elastic modulus parameters based on the characteristic values, and outputs an assessment result including a risk level label and damage location coordinates; wherein the characteristic values ​​include spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor; The control strategy is matched according to the risk level label, and the corresponding device control instructions are generated.

[0006] By adopting the above technical solutions, accurate identification and dynamic early warning of safety risks of large-scale high-altitude steel structure corridors have been achieved, significantly improving the intelligence level and response efficiency of complex structural health monitoring. The solution adopts time-space aligned multi-source data fusion technology, combined with signal processing algorithms such as wavelet packet entropy, empirical mode decomposition and matching pursuit, to effectively solve the problem of severe on-site environmental noise interference and weak and difficult-to-distinguish damage signals; by establishing a high-dimensional damage characteristic tensor containing multi-physical field characteristics such as strain-temperature coupling derivatives and frequency domain energy entropy, it realizes the quantitative characterization of key damage mechanisms such as structural thermal stress effects and fatigue crack propagation; the three-level decision tree risk assessment model constructed based on digital twin technology can achieve real-time response while ensuring calculation accuracy, and ensures that the simulation results are highly consistent with the actual state through local grid refinement and dynamic correction of material parameters; the hierarchical response mechanism ensures the timeliness of risk response and avoids unnecessary waste of resources. This application can effectively deal with the challenges brought by structural complexity, environmental interference and uncertainty, and provide a scientific and reliable monitoring method for the safe operation and maintenance of key infrastructure such as large-scale steel structure corridors.

[0007] Optionally, the step of performing dynamic noise suppression on the sensor data matrix based on a pre-trained environmental noise transfer function model and outputting a noise-reduced pure signal matrix and damage characteristic frequency band identifiers includes: Acquire a time-synchronized sensor data matrix; the sensor data matrix includes vibration signals, acoustic emission signals, wind speed, and temperature and humidity data; Load the pre-trained ambient noise transfer function model; Performing wavelet packet decomposition on the vibration signal and calculating the wavelet packet energy entropy value of each sub-band; According to the energy entropy value, sub-band components with entropy values ​​lower than a set threshold are screened and reconstructed into a primary noise reduction signal; Performing empirical mode decomposition on the primary noise reduction signal to extract intrinsic mode function components; Filtering the intrinsic mode function component related to wind speed based on the environmental noise transfer function model to generate a pure vibration signal; Performing a matching pursuit algorithm on the acoustic emission signal, excluding preset rain noise frequency band atoms when the temperature and humidity data exceeds a preset temperature and humidity threshold, and extracting characteristic atoms that match the crack waveform as crack characteristic waveform atoms; fusing the pure vibration signal with crack characteristic waveform atoms to generate a pure signal matrix; The effective frequency band boundary of the crack characteristic waveform atom is calculated, and a damage characteristic frequency band boundary identifier is output.

[0008] By adopting this technical solution, we achieve precise processing of structural health monitoring signals under complex environmental conditions. This solution not only considers the physical mechanism modeling of environmental interference such as wind-induced noise, but also incorporates advanced signal processing techniques based on information theory and sparse representation. Through multi-dimensional feature extraction and noise suppression, it significantly improves the accuracy and reliability of structural damage identification, providing scientific and reliable technical support for the safety monitoring of large steel structures.

[0009] Optionally, the step of establishing the environmental noise transfer function model includes: Collect historical background noise data during periods without structural loads; Wind speed data were recorded simultaneously as independent variables; Calculate the wind speed autopower spectrum P xx (f) and wind speed-noise cross power spectrum P xy (f); The frequency domain transfer function is calculated by the cross power spectrum density, and the calculation formula is: H(f)=P xy (f) / P xx (f).

[0010] By adopting the above technical solution, a frequency domain model that can accurately describe the wind speed-noise relationship was constructed, providing a reliable theoretical basis and calculation basis for subsequent dynamic noise suppression.

[0011] Optionally, the steps of calculating a thermal stress sensitivity factor based on the strain signal and the temperature signal, extracting multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fusing the features to generate a high-dimensional damage characteristic tensor include: Receive a pure signal matrix and a damage characteristic frequency band identifier; the pure signal matrix includes pure vibration signals and crack characteristic waveform atoms; Synchronously acquire the temporally and spatially aligned raw strain and temperature signals; Based on the strain signal and the temperature signal, solving the partial derivative of strain with respect to temperature within a preset time window to obtain a thermal stress sensitivity factor; Extracting the energy distribution of the crack characteristic waveform atoms in the time domain waveform, and calculating the crack energy entropy value within the damage characteristic frequency band identification range; Performing frequency domain transformation on the pure vibration signal to extract vibration frequency domain features within a preset characteristic frequency band; The thermal stress sensitivity factor, crack energy entropy value and vibration frequency domain characteristics are fused through a 1D-CNN neural network to output a fused high-dimensional damage feature tensor.

[0012] By adopting the above technical solutions, a complete processing chain from multi-physical field signal acquisition to high-dimensional feature fusion was constructed, achieving refined identification and characterization of structural damage. By fully considering multi-dimensional information such as thermal-mechanical coupling effects, time-frequency localization characteristics, and frequency domain dynamic characteristics, a damage identification method with practical engineering value was formed, which improved the accuracy and reliability of the structural health monitoring system.

[0013] Optionally, the high-dimensional damage feature tensor is input into a pre-built digital twin risk assessment model, the digital twin risk assessment model locally refines the finite element mesh according to the feature value and updates the material elastic modulus parameter, and the step of outputting an assessment result including a risk level label and damage location coordinates includes: Receiving the high-dimensional damage feature tensor, including thermal stress sensitivity factor, crack energy entropy value and vibration frequency domain characteristics; Calculating a spatial strain gradient value based on the thermal stress sensitivity factor; When the spatial strain gradient value exceeds a preset threshold, a local refinement operation of the finite element mesh is triggered in the pre-built digital twin risk assessment model; Call the temperature-elastic modulus mapping table in the pre-stored material constitutive relationship library and dynamically update the elastic modulus parameters according to the real-time temperature data; Inputting the crack energy entropy value into a crack growth rate model to calculate a critical damage index; Perform finite element simulation on the refined mesh model after updating the elastic modulus parameters and output the simulated stress field; Based on the preset physical rule threshold, the thermal stress sensitivity factor and the vibration frequency domain characteristics are compared to obtain a first-level decision result; Input the vibration frequency domain features into the pre-trained LSTM time series model, calculate the residual between the predicted value and the actual value, and obtain the secondary decision result; Calculate the modal matching degree between the simulated stress field and the vibration frequency domain characteristics to obtain the three-level decision results; Modifying the preset decision weight ratio according to the critical damage index, and weightedly fusing the first-level decision result, the second-level decision result, and the third-level decision result to generate a risk level label; Extracting peak coordinates in the simulated stress field as damage position coordinates; The risk level label and the damage location coordinates are combined to obtain an assessment result.

[0014] By adopting the above technical solutions, an adaptive, dynamic and high-precision digital twin risk assessment system was constructed. This model can not only achieve comprehensive perception and accurate positioning of structural damage status, but also dynamically adjust the assessment strategy according to different damage development stages, significantly improving the timeliness and accuracy of risk warnings.

[0015] Optionally, after the step of outputting the assessment result including the risk level label and the damage location coordinates, the following steps are further included: Extracting the modal confidence factor deviation value between the measured vibration frequency domain characteristics at the damage location coordinates and the simulated stress field at the same spatial coordinates; When the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times, triggering the constitutive parameter inversion engine; Iteratively optimizing a temperature-elastic modulus mapping table in the material constitutive relationship library based on a covariance matrix adaptive evolutionary algorithm; The optimized temperature-elastic modulus mapping table is updated to the parameter storage area of ​​the digital twin risk assessment model, and a version iteration log is generated.

[0016] By adopting the above technical solutions, a complete closed-loop control system, from model deviation detection to adaptive parameter optimization, was constructed, enabling intelligent maintenance and dynamic updating of the digital twin risk assessment model. The system first quantified the model's prediction accuracy through modal confidence factor deviation analysis. When persistent deviations were detected, the constitutive parameter inversion optimization mechanism based on the CMA-ES algorithm was initiated. Ultimately, precise updates of model parameters were achieved through versioning management, ensuring the digital twin system's ability to maintain long-term stable predictive performance in complex engineering environments and providing reliable technical support for structural health monitoring and fault early warning.

[0017] Secondly, the present application provides a high-altitude large-scale steel structure corridor safety risk monitoring system, which adopts the following technical solutions: A safety risk monitoring system for a large-scale high-altitude steel structure corridor, comprising: A data acquisition module is used to collect raw monitoring data through a multi-source sensor group deployed at each node of the steel structure corridor; the raw monitoring data includes strain signals, vibration signals, acoustic emission signals and temperature signals; A data processing module is used to perform spatiotemporal alignment processing on the raw monitoring data to generate a time-synchronized sensor data matrix; A noise reduction module is used to dynamically suppress the noise of the sensor data matrix based on a pre-trained environmental noise transfer function model, and output a pure signal matrix after noise reduction and a damage characteristic frequency band identifier; a characteristic tensor generation module, configured to calculate a thermal stress sensitivity factor based on the strain signal and the temperature signal, extract multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fuse them to generate a high-dimensional damage characteristic tensor; a risk assessment module, configured to input the high-dimensional damage characteristic tensor into a pre-built digital twin risk assessment model, wherein the digital twin risk assessment model locally refines the finite element mesh and updates the material elastic modulus parameters based on the characteristic values, and outputs an assessment result including a risk level label and damage location coordinates; wherein the characteristic values ​​include spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor; The device control module is used to match the control strategy according to the risk level label and generate corresponding device control instructions.

[0018] Optionally, the monitoring system further includes: A measured data acquisition module, used to obtain the measured vibration frequency domain characteristics at the damage location coordinates; a deviation calculation module, used to calculate the modal confidence factor deviation value between the measured vibration frequency domain characteristics and the simulated stress field at the same spatial coordinates; An over-limit trigger module is used to trigger the constitutive parameter inversion engine when the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times; An iterative optimization module, configured to iteratively optimize a temperature-elastic modulus mapping table in the material constitutive relationship library based on a covariance matrix adaptive evolutionary algorithm; The iteration log generation module is used to update the optimized temperature-elastic modulus mapping table to the parameter storage area of ​​the digital twin risk assessment model and generate a version iteration log.

[0019] In a third aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium stores a computer program capable of being loaded by a processor and executing any one of the methods in the first aspect.

[0020] To sum up, this application includes at least one of the following beneficial technical effects: this application can effectively cope with the challenges brought by structural complexity, environmental interference and uncertainty, and provide a scientific and reliable monitoring method for the safe operation and maintenance of key infrastructure such as large steel structure corridors. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1This is a first flow chart of a method for monitoring safety risks of a large-scale high-altitude steel structure corridor according to one of the embodiments of the present application.

[0022] Figure 2 This is a second flow chart of the method for monitoring safety risks of large-scale high-altitude steel structure corridors in one of the embodiments of the present application.

[0023] Figure 3 This is the third flow chart of the safety risk monitoring method for a large-scale high-altitude steel structure corridor in one of the embodiments of the present application.

[0024] Figure 4 This is the fourth flow chart of the method for monitoring safety risks of large-scale high-altitude steel structure corridors in one of the embodiments of the present application.

[0025] Figure 5 This is the fifth flow chart of the method for monitoring safety risks of large-scale high-altitude steel structure corridors in one of the embodiments of the present application.

[0026] Figure 6 This is the sixth flow chart of the safety risk monitoring method for a large-scale high-altitude steel structure corridor in one of the embodiments of the present application. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figures 1-6 It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0028] The embodiment of the present application discloses a method for monitoring the safety risks of a large-scale high-altitude steel structure corridor.

[0029] Reference Figure 1 A method for monitoring safety risks of a large-scale high-altitude steel structure corridor is provided. The monitoring method includes: Step S101, collecting original monitoring data through a multi-source sensor group deployed at each node of the steel structure corridor; The raw monitoring data includes strain signals, vibration signals, acoustic emission signals, and temperature signals. During the data collection phase, raw monitoring data is acquired through a multi-source sensor set deployed at key nodes of the steel structure corridor (such as support points, connectors, and cantilever ends). These sensors include but are not limited to strain gauges, accelerometers, temperature and humidity sensors, wind speed sensors, and acoustic emission sensors.

[0030] Specifically, strain signals reflect the degree of deformation of a structure under external loads; vibration signals reflect the dynamic response of the structure to environmental excitation or human disturbance; and acoustic emission signals are used to capture the transient elastic wave energy released during the propagation of microcracks within the material. These three types of signals characterize the structural state from different dimensions and form the basic data source for subsequent analysis. This multi-source heterogeneous data acquisition method not only improves the comprehensiveness of monitoring but also provides rich information support for subsequent data fusion and damage identification.

[0031] Step S102, performing spatiotemporal alignment processing on the original monitoring data to generate a time-synchronized sensor data matrix; Among them, since different types of sensors may have different sampling frequencies, transmission delays and timestamp accuracy, if they are not corrected, it will lead to time series misalignment and affect the accuracy of subsequent analysis.

[0032] Therefore, interpolation, resampling, and timestamp alignment can be used to unify the time bases of all sensor channels and construct a unified data matrix structure based on spatial positional relationships. This process essentially creates a four-dimensional data cube: one with a time axis, one with a sensor type axis, one with a spatial node axis, and one with a physical dimension axis. This matrix representation provides a structured input format for subsequent noise suppression and feature extraction, and also facilitates parallel computing and data flow scheduling in subsequent processing modules.

[0033] Step S103: Dynamically suppress the sensor data matrix based on the pre-trained environmental noise transfer function model, and output the clean signal matrix after noise reduction and the damage characteristic frequency band identifier; Among them, the environmental noise transfer function model is usually established through preliminary experiments or historical data analysis to describe the propagation path of environmental noise in a specific structure and its impact mechanism on each sensor signal.

[0034] Based on this, differentiated noise reduction strategies are adopted for different signal types. For vibration signals, a hybrid noise reduction algorithm is developed, combining wavelet packet entropy analysis with empirical mode decomposition (EMD). The wavelet packet transform decomposes the signal into multiple frequency bands, identifying the main energy distribution areas by calculating the wavelet packet entropy of each subband, thereby preserving the effective components and filtering out redundant noise. EMD adaptively decomposes non-stationary signals into several intrinsic mode functions (IMFs), and then further filters out the true vibration components through the Hilbert transform. The combination of these two methods enhances the signal-to-noise ratio while preserving the original vibration characteristics.

[0035] For acoustic emission signals, a matching pursuit algorithm is used to isolate characteristic frequency bands related to crack propagation. Matching pursuit is a sparse approximation method based on an atom library. By iteratively selecting optimal atoms and matching them with the signal residuals, it gradually reconstructs the main components containing information about crack activity and determines the corresponding set of frequency boundary values, known as the damage characteristic frequency band identifier. This identifier serves as a key parameter for subsequent feature extraction, limiting the analysis scope and improving detection sensitivity.

[0036] Step S104: Calculate the thermal stress sensitivity factor based on the strain signal and the temperature signal, extract the multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fuse them to generate a high-dimensional damage characteristic tensor; Specifically, this stage aims to mine high-dimensional features closely related to structural damage from the pure signal matrix to form a high-dimensional damage feature tensor that can be used for risk assessment.

[0037] First, the partial derivative of the strain data with respect to temperature changes is calculated to obtain the thermal stress sensitivity factor. Steel structures experience significant thermal expansion and contraction due to diurnal temperature fluctuations or climate change, which in turn induces additional stresses. By solving the partial differential relationship between strain and temperature, the impact of thermal stress on structural stability can be quantified. This is particularly useful for identifying problems such as local buckling or loose connections caused by excessive temperature gradients.

[0038] It should be noted that dynamic noise suppression is not performed on the strain signal at the front end because its core interference is temperature drift, not environmental noise. This interference is eliminated through subsequent coupled derivative calculations driven by the physical model. However, vibration and acoustic emission signals require dedicated suppression algorithms due to frequency domain aliasing and transient characteristic volatility. This layered processing strategy improves data processing efficiency.

[0039] Secondly, a frequency domain transform is performed on the pure vibration signal to extract the vibration frequency domain characteristics within a preset characteristic frequency band. Simultaneously, the crack energy entropy value (used to characterize the intensity of crack activity) of the acoustic emission signal within the damage characteristic frequency band is extracted. Energy entropy reflects the uniformity of the signal's energy distribution within a specified frequency band. If the energy in a certain frequency band is concentrated and continuously rising, it often indicates that the crack is expanding or fatigue damage is intensifying.

[0040] Finally, the vibration frequency domain features, thermal stress sensitivity factors, and energy entropy values ​​are fused through a 1D-CNN neural network. 1D-CNN has excellent time series modeling capabilities and can automatically extract periodic patterns, resonance peak positions, and amplitude variation trends from vibration signals. These are then combined with thermal stress factors and energy entropy to form a high-dimensional nonlinear mapping relationship, ultimately outputting a high-dimensional damage feature tensor.

[0041] Step S105: Input the high-dimensional damage feature tensor into a pre-built digital twin risk assessment model. The digital twin risk assessment model locally refines the finite element mesh based on the feature value and updates the material elastic modulus parameters, outputting an assessment result including a risk level label and damage location coordinates. Among them, the eigenvalues ​​include the spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor; In the embodiment of the present application, the digital twin risk assessment model integrates finite element simulation, real-time data update and multi-level decision-making mechanism. When the strain gradient in the characteristic tensor exceeds the preset threshold, the system automatically triggers the local mesh refinement operation, and adjusts the finite element mesh size of the corresponding area to 0.1mm level to improve the calculation accuracy of the area and more accurately simulate the potential damage evolution process. At the same time, considering the significant effect of temperature on material properties, the model also dynamically corrects the calculation formula parameters of the material elastic modulus according to the real-time temperature data to ensure that the simulation results are consistent with the actual working conditions.

[0042] In terms of assessment logic, the model uses a three-level decision tree structure to determine risk levels: the first level uses physical rule thresholds to determine whether indicators such as strain and frequency drift exceed regulatory limits; the second level uses an LSTM neural network to predict historical trends and analyze the residual between the current state and the predicted value. If the residual continues to increase, it indicates that the structural state has deviated from its normal trajectory; the third level evaluates the consistency of the overall structural behavior by comparing the deviation between the current state and the simulated output of the digital twin risk assessment model. This multi-level, multi-dimensional assessment mechanism effectively improves the robustness and reliability of risk identification.

[0043] Step S106: Match the control strategy according to the risk level label and generate corresponding equipment control instructions.

[0044] Among them, the design of the regulatory strategy fully considers the differences in risk levels and the timeliness of emergency response.

[0045] For example, when the risk level is "level one", it means that the structure is in an early abnormal state and has not yet reached a significant dangerous level. At this time, by increasing the sampling frequency of the sensors around the damage location to 1kHz, it is possible to capture the abnormal signals in a refined manner, providing more detailed information for subsequent diagnosis. When the risk level rises to "level two", it indicates that the damage has a tendency to develop further. At this time, a microscopic thermal imager re-inspection instruction containing the coordinates of the damage location is sent to the drone dispatch system. With the help of the high-resolution imaging equipment carried by the drone, a visual inspection of the suspicious area is carried out to help confirm the damage shape and development level. When the risk level reaches "level three", it means that the structure is in a high-risk state and the emergency plan must be activated immediately, including triggering the sound and light alarms at the entrances and exits of the corridor to warn people to evacuate, and simultaneously activating the flow-limiting gate to prevent more people from entering the dangerous area.

[0046] It is understandable that this hierarchical response mechanism not only ensures the timeliness of risk response, but also avoids unnecessary waste of resources.

[0047] In the above implementation, accurate identification and dynamic early warning of safety risks of large-scale high-altitude steel structure corridors are achieved, which significantly improves the intelligence level and response efficiency of complex structural health monitoring. The solution adopts time-space aligned multi-source data fusion technology, combined with signal processing algorithms such as wavelet packet entropy, empirical mode decomposition and matching pursuit, to effectively solve the problem of severe on-site environmental noise interference and weak and difficult-to-distinguish damage signals; by establishing a high-dimensional damage characteristic tensor containing multi-physical field characteristics such as strain-temperature coupling derivatives and frequency domain energy entropy, it realizes the quantitative characterization of key damage mechanisms such as structural thermal stress effects and fatigue crack propagation; the three-level decision tree risk assessment model constructed based on digital twin technology can achieve real-time response while ensuring calculation accuracy, and ensures that the simulation results are highly consistent with the actual state through local grid refinement and dynamic correction of material parameters; the hierarchical response mechanism ensures the timeliness of risk response and avoids unnecessary waste of resources. The present application can effectively respond to the challenges brought by structural complexity, environmental interference and uncertainty, and provide a scientific and reliable monitoring method for the safe operation and maintenance of key infrastructure such as large-scale steel structure corridors.

[0048] Reference Figure 2 As an implementation method of step S103, the steps of performing dynamic noise suppression on the sensor data matrix based on the pre-trained environmental noise transfer function model and outputting the noise-reduced pure signal matrix and damage characteristic frequency band identifiers include: Step S201, obtaining a time-synchronized sensor data matrix; wherein the sensor data matrix includes vibration signals, acoustic emission signals, wind speed, and temperature and humidity data; Time synchronization is a fundamental prerequisite for multi-physics coupling analysis. Due to differences in sampling frequencies and response characteristics among different sensors, the lack of precise time alignment will lead to deviations in subsequent feature extraction based on phase and causal relationships. Within the sensor data matrix, vibration signals reflect changes in the overall dynamic characteristics of the structure, acoustic emission signals capture transient information on the initiation and propagation of microscopic cracks within the material, and wind speed, temperature, and humidity data provide the necessary boundary conditions for environmental load modeling and noise source identification.

[0049] Step S202, loading a pre-trained environmental noise transfer function model; Among them, the environmental noise transfer function model constructs a frequency domain transfer relationship with wind speed as the independent variable. This design is based on the generation mechanism of wind-induced noise in fluid mechanics, that is, the vortex shedding and turbulent pulsation generated when the wind flows through the surface of the structure will stimulate the structure to produce a vibration response with specific frequency components.

[0050] In some possible implementations, the environmental noise transfer function model is established by: collecting background noise samples during a period without structural load; synchronously recording wind speed data as an independent variable; and calculating the wind speed autopower spectrum P. xx (f) and wind speed-noise cross power spectrum P xy (f); calculate the frequency domain transfer function by cross power spectrum density: H(f)=P xy (f) / P xx (f).

[0051] It's clear that by collecting background noise samples during periods without structural load and simultaneously recording wind speed data, a wind speed-noise frequency domain mapping relationship was established. This modeling approach fully utilizes the "quiet period" information of the structure when it's not operating, avoiding interference from the structure's true response signals in noise modeling. The cross-power spectral density calculation method effectively extracts the linear influence of wind speed variations on the noise spectrum distribution, providing a theoretical basis for subsequent adaptive noise suppression.

[0052] Step S203, performing wavelet packet decomposition on the vibration signal and calculating the wavelet packet energy entropy value of each sub-band; Among them, wavelet packet decomposition has a more flexible frequency band division capability than traditional wavelet decomposition, and can achieve uniform division of the signal spectrum, with each sub-band corresponding to a specific frequency interval.

[0053] In some embodiments, energy entropy is an important indicator for measuring signal complexity and information content, and its calculation formula is E i =-∑P i *log(P i ), where P i It represents the ratio of the energy of the i-th sub-band to the total energy.

[0054] Step S204, screening sub-band components with entropy values ​​lower than a set threshold according to the energy entropy values, and reconstructing them into a primary noise reduction signal; Specifically, in structural health monitoring, environmental noise typically manifests as a broadband random signal with relatively uniform energy distribution across subbands, resulting in high energy entropy. In contrast, characteristic signals generated by structural damage are often concentrated in specific frequency bands, with uneven energy distribution and relatively low entropy. By setting a threshold to filter out low-entropy subbands, we can effectively identify frequency bands containing the true structural response information. This information-theoretic feature extraction method exhibits excellent noise robustness.

[0055] Step S205, performing empirical mode decomposition on the primary noise reduction signal to extract intrinsic mode function components; The core of empirical mode decomposition (EMD) lies in decomposing complex nonlinear and nonstationary signals into several intrinsic mode function (IMF) components. Each IMF component meets two conditions: the number of extreme points and the number of zero crossings are equal or differ by no more than 1; and the mean of the upper and lower envelopes is zero. This adaptive decomposition method constructs basis functions based on the time-frequency characteristics of the signal itself, avoiding the spectral leakage problem that may be caused by pre-set basis functions.

[0056] Step S206: filtering out the intrinsic mode function component related to wind speed based on the environmental noise transfer function model to generate a pure vibration signal; Among them, after obtaining the IMF components, the components related to wind speed are filtered out based on the pre-trained environmental noise transfer function model. This operation is essentially to fuse the time-frequency analysis results of empirical mode decomposition with the frequency domain modeling results of the transfer function. By identifying the main frequency components of each IMF component, the transfer function model is used to determine whether it belongs to the category of wind-induced noise, thereby achieving targeted noise suppression.

[0057] Step S207: performing a matching pursuit algorithm on the acoustic emission signal, excluding the preset rain noise frequency band atoms when the temperature and humidity data exceed the preset temperature and humidity threshold, and extracting the characteristic atoms that match the crack waveform as the crack characteristic waveform atoms; Among them, the matching pursuit algorithm is based on the concept of overcomplete dictionary and realizes sparse representation of the signal by iteratively selecting the atoms that best match the signal.

[0058] For example, when humidity > 90% and temperature > 5°C, a rainfall environment is determined. At this point, atoms in a preset rain noise frequency band (e.g., atoms in the 8-10 kHz band) are removed from the overcomplete dictionary. This process is based on the spectral characteristics of rainfall noise: raindrops impacting sensors or structural surfaces produce an impact response within a specific frequency range. Simultaneously, the atomic matching calculation focuses on characteristic atoms that match the crack waveform (e.g., the 30-150 kHz band). This frequency band selection is based on the propagation characteristics of elastic waves generated during crack growth in metal materials. Stress release at the crack tip generates high-frequency stress waves, typically ranging from tens to hundreds of kHz.

[0059] Step S208, fusing the pure vibration signal with the crack characteristic waveform atoms to generate a pure signal matrix; Among them, a complete structural state representation matrix is ​​constructed by integrating the characteristic signals obtained from different processing paths.

[0060] Step S209: Calculate the effective frequency band boundary of the crack characteristic waveform atom and output the damage characteristic frequency band boundary identifier.

[0061] Among them, the center frequency f of the crack characteristic waveform atom is extracted c and bandwidth Bw ; Calculate the effective frequency band boundary: [f c −αB w ,f c +αB w ], where α is the preset bandwidth expansion coefficient. This boundary calculation method takes into account the spectrum diffusion characteristics of the actual signal. By adjusting the bandwidth expansion coefficient, it can ensure the integrity of feature extraction while avoiding noise mixing caused by excessive frequency band.

[0062] The above implementation achieves precise processing of structural health monitoring signals under complex environmental conditions. This solution not only considers the physical mechanism modeling of environmental interference such as wind-induced noise, but also incorporates advanced signal processing techniques based on information theory and sparse representation. Through multi-dimensional feature extraction and noise suppression, it significantly improves the accuracy and reliability of structural damage identification, providing scientific and reliable technical support for the safety monitoring of large steel structures.

[0063] Reference Figure 3 As an implementation of the environmental noise transfer function model, the steps of establishing the environmental noise transfer function model include: Step S301, collecting background noise data during a historical period without structural load; The unloaded period refers to the time window when the structure is inactive or not subject to significant loads. During this period, the effective response signal generated by the structure itself is negligible, and the signal collected by the sensor is primarily composed of ambient noise. This data acquisition strategy embodies the core concept of "signal-noise separation" in modern signal processing. By selecting an appropriate observation window to obtain pure noise samples, it avoids interference from the structure's true response signal on noise feature extraction.

[0064] Step S302, synchronously record wind speed data as an independent variable; Wind speed is the primary driver of environmental noise, and its interaction with structural surfaces, generating fluid dynamics phenomena such as eddies and turbulence, is a significant source of environmental noise. By synchronously recording wind speed data, a temporal relationship between environmental excitation and noise response is established. This synchronization ensures the accuracy of causal relationships in subsequent transfer function calculations.

[0065] Step S303: Calculate the wind speed autopower spectrum P xx (f) and wind speed-noise cross power spectrum P xy (f); Among them, the autopower spectrum P xx (f) describes the energy distribution of the wind speed signal x(t) at each frequency component. Its essence is the modulus square of the Fourier transform of the wind speed signal, reflecting the frequency domain characteristics of wind speed changes. The specific calculation formula is: ; In addition, the cross power spectrum P xy (f) describes the frequency domain correlation between the wind speed signal x(t) and the noise signal y(t), and includes information about the amplitude and phase relationships of the two signals at each frequency component. The calculation of the cross-power spectrum requires a cross-correlation operation on the two signals followed by a Fourier transform. The result is a complex number, the real part of which reflects the relationship between the in-phase components of the two signals, and the imaginary part reflects the relationship between the orthogonal components. The specific calculation formula is: ; In the above formula, the Fourier transform of wind speed signal x(t) and noise signal y(t) are X(f) and Y(f), respectively. represents the conjugate complex number of X(f), and T represents the time length of signal observation.

[0066] In the embodiments of the present application, these power spectrum calculations adopt the classic Welch method or periodogram method, and improve the stability of spectrum estimation through segmented averaging, effectively reducing the impact of random fluctuations on the results.

[0067] Step S304: Calculate the frequency domain transfer function by using the cross power spectrum density. The calculation formula is: H(f)=P xy (f) / P xx (f).

[0068] The transfer function mathematically describes the system's response to different frequency components. Its numerator, Pxy(f), contains information about the coupling between the input and output signals, while the denominator, Pxx(f), normalizes the result, eliminating the influence of variations in the input signal's intensity. The modulus of the transfer function reflects the system's amplification or attenuation of each frequency component, while the phase describes the time delay characteristics of each frequency component.

[0069] In the embodiment of the present application, the transfer function substantially establishes a quantitative relationship between wind speed changes and the spectral distribution of ambient noise, providing an accurate mathematical model support for subsequent adaptive noise suppression.

[0070] In the above implementation, a frequency domain model that can accurately describe the wind speed-noise relationship is constructed, providing a reliable theoretical basis and calculation basis for subsequent dynamic noise suppression.

[0071] Reference Figure 4 As an implementation method of step S104, the steps of calculating the thermal stress sensitivity factor based on the strain signal and the temperature signal, extracting the multi-physics field coupling characteristics associated with the damage characteristic frequency band identifier in the pure signal matrix, and fusing them to generate a high-dimensional damage characteristic tensor include: Step S401, receiving a clean signal matrix and an impairment characteristic frequency band identifier; The clean signal matrix includes clean vibration signals and crack signature waveform atoms. The clean vibration signal refers to the true structural response signal obtained after preliminary noise suppression processing, while the crack signature waveform atoms are basis functions with specific time-frequency localization characteristics extracted using matching pursuit algorithms or dictionary learning techniques. They can effectively characterize the transient characteristics of structural damage.

[0072] Step S402, synchronously acquiring the original strain signal and temperature signal aligned in time and space; The spatiotemporal alignment of strain and temperature signals requires that the two sensors be placed in representative physical locations. These sensors are typically placed at the same or adjacent measuring points to ensure that the acquired physical quantities reflect the true state of the local area of ​​the structure. Time synchronization is achieved through hardware clock synchronization or software interpolation algorithms. Timestamp alignment accuracy is typically required to be in the millisecond range or higher to ensure the accuracy of subsequent differential operations.

[0073] Step S403: Based on the strain signal and the temperature signal, the partial derivative of the strain with respect to the temperature is solved within a preset time window to obtain a thermal stress sensitivity factor; The thermal stress sensitivity factor essentially reflects the sensitivity of a material or structure to stress-temperature coupling. Its physical significance lies in describing the change in strain caused by a unit temperature change, making it a key parameter in thermal stress analysis. The sliding time window approach takes into account the time-varying nature of the structural response. The selection of the window length requires a trade-off between temporal resolution and statistical stability, and is typically determined based on the characteristic period of the signal and the sampling frequency.

[0074] Specifically, the calculation of the thermal stress sensitivity factor involves constructing a sliding time window with the temperature signal as the independent variable and the strain signal as the dependent variable. The linear relationship between strain and temperature is fitted using the least squares method, with the slope being the partial derivative. This method obtains optimal parameter estimates by minimizing the sum of squared residuals between the observed data and the theoretical model, demonstrating excellent numerical stability and statistical properties. The calculation of the slope, as an estimate of the partial derivative, involves matrix operations and linear algebra theory, specifically the solution of normal equations.

[0075] Step S404, extracting the energy distribution of the crack characteristic waveform atoms in the time domain waveform, and calculating the crack energy entropy value within the damage characteristic frequency band identification range; The energy entropy calculation of the crack characteristic waveform atom includes: analyzing the time domain envelope of the atomic waveform function gγ(t); integrating the signal energy E in the damage characteristic frequency band identification range; i According to the entropy formula H=−∑(E i / E total)log2(E i / E total ) to calculate the entropy value.

[0076] Specifically, the time-domain envelope analysis of crack characteristic waveform atoms uses the Hilbert transform technique, which extracts instantaneous amplitude information by converting real signals into analytical signals. The envelope function can effectively reflect the amplitude modulation characteristics of the signal, which is of great significance for identifying the transient impact response generated by cracks. The integral calculation of energy distribution is essentially a cumulative measure of signal power within a specific frequency band, reflecting the application of Parseval's theorem within a finite frequency band. The degree of energy concentration of the signal within the damage-sensitive frequency band is quantified through frequency domain integration. The entropy value calculation uses the Shannon entropy formula to measure the uncertainty or complexity of the signal energy distribution. When the crack develops and the signal characteristics change, the entropy value of the energy distribution will also change accordingly, thereby providing a quantitative indicator for damage identification.

[0077] Step S405: performing frequency domain transformation on the pure vibration signal to extract the vibration frequency domain features within a preset characteristic frequency band; Frequency domain transformation typically uses the Fast Fourier Transform (FFT) algorithm to convert time domain signals into a frequency domain representation. This transform, based on Fourier analysis theory, can decompose complex time domain waveforms into a superposition of different frequency components. Frequency domain features can be extracted in various forms, including amplitude spectrum, phase spectrum, and power spectral density. These features can reflect the dynamic characteristics of the structure and provide a frequency domain basis for damage identification.

[0078] In some embodiments, the preset characteristic frequency band is 0-200 Hz, which is suitable for steel structure corridors with a span of 30-80 meters. The selection of the preset characteristic frequency band is based on structural dynamics theory and modal analysis results. The fundamental frequency of large steel structures is generally distributed within this range.

[0079] Step S406: The thermal stress sensitivity factor, the crack energy entropy value, and the vibration frequency domain characteristics are fused through a 1D-CNN neural network to output a fused high-dimensional damage feature tensor.

[0080] The design of the 1D-CNN neural network architecture leverages the temporal characteristics of one-dimensional signals, automatically extracting and fusing multidimensional features through a combination of convolutional, pooling, and fully connected layers. Thermal stress sensitivity factors provide information on thermal-mechanical coupling, crack energy entropy reflects the time-frequency localization of damage, and vibration frequency domain features reflect the overall dynamic characteristics of the structure. These three types of features complement each other in a physical sense, forming a multidimensional damage characterization system. The resulting high-dimensional damage feature tensor is rich in semantic information and has excellent discriminative power, providing high-quality feature representation for subsequent damage localization and quantitative assessment.

[0081] In some embodiments, the training process of the 1D-CNN neural network is based on the back-propagation algorithm and gradient descent optimization theory, and a nonlinear mapping relationship between input features and damage status is established through learning from a large number of samples.

[0082] In the above implementation, a complete processing chain from multi-physical field signal acquisition to high-dimensional feature fusion is constructed, which realizes the refined identification and characterization of structural damage. By fully considering multi-dimensional information such as thermal-mechanical coupling effects, time-frequency localization characteristics and frequency domain dynamic characteristics, a damage identification method with practical engineering value is formed, which improves the accuracy and reliability of the structural health monitoring system.

[0083] Reference Figure 5 As an implementation method of step S105, the high-dimensional damage feature tensor is input into a pre-built digital twin risk assessment model. The digital twin risk assessment model locally refines the finite element mesh according to the feature value and updates the material elastic modulus parameter. The steps of outputting the assessment result including the risk level label and the damage location coordinates include: Step S501, receiving a high-dimensional damage feature tensor, including a thermal stress sensitivity factor, a crack energy entropy value, and a vibration frequency domain feature; Step S502, calculating the spatial strain gradient value based on the thermal stress sensitivity factor; The calculation of spatial strain gradients essentially quantifies the rate of change of thermal stress-sensitive factors within the spatial domain. This process, based on partial differential equation theory, approximates the spatial derivative of the strain field using finite difference or finite element methods. The physical significance of the gradient value lies in describing the intensity of spatial changes in the strain field. Regions of high gradient typically correspond to locations of stress concentration or material discontinuities, and these areas are often prone to structural damage.

[0084] Step S503: When the spatial strain gradient value exceeds a preset threshold, a local refinement operation of the finite element mesh is triggered in the pre-built digital twin risk assessment model; It's important to note that traditional finite element simulations often use a uniform meshing scheme, making it difficult to balance global accuracy and computational efficiency. In this solution, the system monitors changes in strain gradients in real time and dynamically adjusts the mesh density in key areas of the model, thereby improving simulation accuracy in high-risk areas without significantly increasing the overall computational burden.

[0085] This adaptive mesh refinement strategy based on physical field gradients is widely used in fields such as structural mechanics and fluid mechanics. Its advantage is that it can significantly improve the ability to depict local details while maintaining computational efficiency. It is particularly suitable for dealing with damage evolution problems with strong non-uniformity and multi-scale characteristics.

[0086] Step S504: calling a temperature-elastic modulus mapping table in a pre-stored material constitutive relationship library, and dynamically updating the elastic modulus parameters according to real-time temperature data; Among them, the elastic modulus of a material, an important parameter that describes its ability to resist elastic deformation, will change significantly in a high-temperature environment. If the material parameters at room temperature are still used, the simulation results will be distorted. To this end, the system dynamically adjusts the material properties of the corresponding units in the model by accessing the elastic modulus curves or functional relationships corresponding to different temperature conditions stored in the material database, combining them with the actual temperature data currently collected. This parameter update mechanism driven by measured data not only improves the accuracy of the model, but also enhances its adaptability to real working conditions.

[0087] Step S505, inputting the crack energy entropy value into the crack growth rate model to calculate the critical damage index; Specifically, the system uses fracture mechanics theory to establish a functional relationship between crack growth rate, external load, and material properties. It then inverts the crack development trend based on the current crack energy entropy value and calculates a critical damage index that comprehensively reflects the structural safety margin. This index can be considered a warning threshold before the structure enters a dangerous state; the closer its value is to 1, the closer the structure is to the critical point of failure. By introducing this index, the system can rationally allocate the weights of different assessment sub-models in the subsequent decision-making process, achieving dynamic optimization of risk level judgment.

[0088] Step S506, performing finite element simulation on the refined mesh model after updating the elastic modulus parameters, and outputting a simulated stress field; Specifically, through the aforementioned adaptive mesh refinement and material parameter updates, the system constructs a refined finite element model that more closely reflects actual operating conditions. Structural mechanics simulations based on this model can more accurately simulate the stress distribution of the structure under complex loads. The resulting stress field not only captures the overall stress state of the structure but also reveals the location and strength of localized high-stress areas, providing a crucial basis for subsequent damage location and risk assessment.

[0089] Step S507: Based on the preset physical rule threshold, the thermal stress sensitivity factor and the vibration frequency domain characteristics are compared to obtain a first-level decision result; The system uses physical rules based on expert experience or experimental data to perform threshold comparisons on both types of features. For example, when the thermal stress sensitivity factor exceeds a certain critical value, the system determines that the area may be at risk of thermal fatigue damage; when a certain modal frequency in the vibration frequency domain falls below a preset lower limit, it may indicate a decrease in structural stiffness. This preliminary judgment based on physical laws is highly interpretable and robust, and can provide a basic basis for risk assessment even in the absence of a large number of training samples.

[0090] Step S508: Input the vibration frequency domain features into the pre-trained LSTM time series model, calculate the residual between the predicted value and the actual value, and obtain the secondary decision result; LSTM (Long Short-Term Memory) networks, as a specialized recursive neural network structure, excel at capturing long-term dependencies in time series data. By pre-training the model on historical data, it learns how structural responses evolve over time. During operation, the system inputs the currently collected vibration frequency characteristics into the model to obtain a predicted output for the future state. The system then calculates the residual between the predicted and measured values ​​to determine whether the structure is exhibiting abnormal behavior. A larger residual indicates a more severe deviation from normal behavior and a higher potential risk. This approach overcomes the limitations of static threshold judgments and enhances the ability to perceive dynamic evolutionary processes.

[0091] Step S509, calculating the modal matching degree between the simulated stress field and the vibration frequency domain characteristics to obtain a three-level decision result; Using modal analysis theory, the system compares simulated stress field information with measured vibration frequency domain characteristics, assessing the degree of similarity between the two in terms of frequency, mode shape, and other aspects. A higher degree of modal matching indicates that the simulated model more accurately reflects the structure's true dynamic characteristics; a lower degree of matching may indicate model deviations or structural damage. This step, by incorporating modal consistency analysis, enhances the objectivity and reliability of the evaluation results and avoids the risk of misjudgment due to a single feature source.

[0092] Step S510: modifying the preset decision weight ratio according to the critical damage index, and performing weighted fusion on the first-level decision results, the second-level decision results, and the third-level decision results to generate a risk level label; Specifically, since different types of risk assessment methods have their own advantages and disadvantages, relying solely on one method may lead to misjudgment. Therefore, the system dynamically adjusts the weight distribution of each sub-model based on the critical damage index output by the crack growth rate model. For example, when the critical damage index is low, it indicates that the structure is still in the early damage stage. At this time, the first-level assessment based on physical rules is more reliable and is therefore given a higher weight; as the damage intensifies, the importance of the time series model and modal matching analysis increases, and the corresponding weights are adjusted accordingly. This weight adaptive mechanism based on the damage evolution state significantly improves the accuracy and adaptability of risk assessment.

[0093] For example, the weight ratio is dynamically allocated according to the critical damage index Dc. When Dc<0.3, the weight ratio of the first-level, second-level, and third-level decision results is configured as 4:3:3; when 0.3≤Dc<0.7, the weight ratio is configured as 5:3:2; when Dc≥0.7, the weight ratio is configured as 6:2:2.

[0094] Step S511, extracting the peak coordinates in the simulated stress field as the damage position coordinates; The system post-processes the simulation results, identifies the maximum points in the stress concentration area, and maps them to a three-dimensional coordinate system, thereby accurately locating the specific location of potential damage. This method combines the high-resolution advantages of numerical simulation with the engineering requirements of damage localization, providing an intuitive spatial reference for subsequent maintenance decisions.

[0095] Step S512: combining the risk level label and the damage location coordinates to obtain an assessment result.

[0096] In the above implementation, by constructing a digital twin risk assessment system with adaptability, dynamics and high precision, the model can not only achieve comprehensive perception and precise positioning of the structural damage status, but also dynamically adjust the assessment strategy according to different damage development stages, significantly improving the timeliness and accuracy of risk warnings.

[0097] In the embodiments of this application, the digital twin risk assessment model employs a hierarchical and progressive architecture, achieving multi-scale quantitative assessment of structural health through a deep fusion of physical mechanisms and data-driven approaches. The model's core input is a three-dimensional, high-dimensional damage characteristic tensor, which includes three key physical parameters: thermal stress sensitivity factor, crack energy entropy, and vibration frequency domain characteristics. In the initial processing phase, the system calculates spatial strain gradients based on the thermal stress sensitivity factor. This process quantifies the local rate of change of the strain field using a spatial difference algorithm. When the gradient exceeds a preset threshold (e.g., 5.0 MPa / °C·m⁻¹), the digital twin risk assessment model's local mesh adaptive optimization mechanism is automatically triggered. This mechanism generates a spherical mesh domain in the potential damage area, whose radius is topologically correlated with the distribution of adjacent sensors. Micron-level refined elements are used within the mesh domain, and a gradient mesh transition strategy is implemented in the transition zone. This significantly reduces the overall computational load while ensuring high-resolution simulation of stress concentration areas.

[0098] At the same time, the model synchronously calls upon a pre-stored library of material constitutive relations, dynamically interpolating and updating the elastic modulus parameters based on real-time ambient temperature data. This process incorporates temperature-modulus attenuation curves for 12 types of structural steel, accurately reflecting the nonlinear attenuation characteristics of material stiffness under high-temperature environments through a quadratic function model. After completing mesh topology optimization and updating material parameters, the system inputs the crack energy entropy value into the Paris crack growth rate model and calculates the critical damage index (ranging from 0 to 1) by inverting the crack evolution trajectory. This index serves as a core indicator for quantifying structural safety margins and directly influences subsequent decision-making weight allocation strategies.

[0099] At the refined simulation level, the model performs coupled thermal-mechanical finite element calculations, outputting a high-resolution three-dimensional stress field distribution. This stress field carries spatial coordinate information, and its peak coordinates are directly mapped to the damage location output. The decision-making layer employs a triple verification mechanism: the first-level decision compares the degree of deviation between the thermal stress sensitivity factor and the vibration fundamental frequency based on a physical rule threshold; the second-level decision analyzes the time series residuals of the vibration frequency domain characteristics through a pre-trained bidirectional LSTM network to capture dynamic characteristic anomalies; and the third-level decision calculates the confidence factor between the simulated stress field and the measured vibration mode to assess the dynamic consistency between the model and the entity. The three decision results are fused via a dynamic weight allocator, whose weight configuration strictly adheres to the segmentation rules guided by the critical damage index. The final fusion result is input into the risk assessment engine, which generates three-level risk level labels based on pre-configured threshold intervals. These labels, combined with the spatial coordinates of the damage extracted from the stress field, form the assessment output.

[0100] In summary, the digital twin risk assessment model constructed in this application solves the three major technical bottlenecks in traditional methods, namely, imbalance between simulation accuracy and efficiency, neglect of temperature effects, and misjudgment of damage stages, through a four-fold closed loop of strain gradient perception, dynamic parameter update, damage entropy feedback, and decision weight adaptation.

[0101] Reference Figure 6 As a further embodiment of the risk monitoring method, after the step of outputting the assessment result including the risk level label and the damage location coordinates, the method further includes: Step S601, extracting the modal confidence factor deviation value between the measured vibration frequency domain characteristics and the simulated stress field at the damage location coordinates in the same spatial coordinates; The modal confidence factor (MCF) is a dimensionless metric used to quantify the correlation between two modal vectors. It is obtained by normalizing the dot product of the vibration mode vectors. By calculating the deviation between the measured and simulated MCFs at the damage location coordinates, the system can quantify the degree of difference between the model's prediction accuracy and the actual measurement results. This deviation reflects the credibility of the digital twin risk assessment model under the current parameter configuration.

[0102] Step S602: When the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times, triggering the constitutive parameter inversion engine; In one of the embodiments of the present application, when the calculated modal confidence factor deviation value exceeds the preset tolerance threshold for three consecutive times, the system will trigger the constitutive parameter inversion engine. The setting of the tolerance threshold needs to comprehensively consider the accuracy limitations of the measurement system, environmental noise interference, and the actual needs of engineering applications. It is usually determined by statistical methods such as the 3σ criterion or empirical distribution based on historical data. The triggering mechanism of the constitutive parameter inversion engine embodies the core idea of ​​the adaptive control system, that is, by monitoring the deviation between the system output and the expected value in real time, the parameter adjustment mechanism is automatically started when the deviation exceeds the acceptable range. This design ensures that the digital twin risk assessment model can dynamically adapt to changes in actual working conditions and maintain the accuracy of the model prediction.

[0103] Step S603, iteratively optimizing the temperature-elastic modulus mapping table in the material constitutive relationship library based on the covariance matrix adaptive evolutionary algorithm; The core advantage of the covariance matrix adaptive evolutionary algorithm lies in its ability to adaptively adjust the covariance matrix of the search distribution, thereby achieving efficient global optimization in high-dimensional parameter spaces. The algorithm generates candidate solutions by maintaining a multivariate Gaussian distribution. As the iteration process progresses, the mean of the distribution shifts toward more optimal solutions, and the covariance matrix is ​​adaptively updated based on the distribution information of successful individuals, allowing the search process to gradually converge to the optimal solution while maintaining exploration capabilities.

[0104] In the embodiment of the present application, the temperature-elastic modulus mapping table is used as the optimization target, and its parameter space usually has high-dimensional nonlinear characteristics. The traditional gradient optimization method is prone to fall into local optimality. The CMA-ES algorithm can effectively overcome this problem through the population evolution mechanism and adaptive adjustment of the covariance matrix, and realize the accurate identification of the material constitutive parameters.

[0105] Step S604: Update the optimized temperature-elastic modulus mapping table to the parameter storage area of ​​the digital twin risk assessment model and generate a version iteration log.

[0106] The temperature-elastic modulus mapping table optimized by the adaptive evolutionary algorithm is updated to the parameter storage area of ​​the digital twin risk assessment model, and a corresponding version iteration log is generated. This version iteration log not only records the timestamp and specific numerical changes of the parameter updates, but also contains key statistical information from the optimization process, such as the objective function convergence curve, number of iterations, and computational time. This information provides an important basis for subsequent model validation, parameter backtracking, and system performance analysis. This mechanism also ensures that the digital twin risk assessment model can continuously learn and improve itself, maintaining its predictive accuracy and decision-making reliability in the face of complex and changing engineering environments.

[0107] In this implementation, a complete closed-loop control system, from model deviation detection to adaptive parameter optimization, is constructed, enabling intelligent maintenance and dynamic updating of digital twin risk assessment models. The system first quantifies model prediction accuracy through modal confidence factor deviation analysis. When persistent deviations are detected, a constitutive parameter inversion optimization mechanism based on the CMA-ES algorithm is initiated. Ultimately, precise model parameter updates are achieved through versioning management. This ensures the digital twin system maintains long-term stable predictive performance in complex engineering environments, providing reliable technical support for structural health monitoring and fault early warning.

[0108] The embodiment of the present application also discloses a safety risk monitoring system for a large-scale high-altitude steel structure corridor.

[0109] A safety risk monitoring system for a large-scale high-altitude steel structure corridor, the monitoring system includes: The data acquisition module is used to collect raw monitoring data through a multi-source sensor group deployed at each node of the steel structure corridor; the raw monitoring data includes strain signals, vibration signals, acoustic emission signals and temperature signals; The data processing module is used to perform spatiotemporal alignment processing on the original monitoring data to generate a time-synchronized sensor data matrix; The noise reduction module is used to dynamically suppress the noise of the sensor data matrix based on the pre-trained environmental noise transfer function model, and output the pure signal matrix after noise reduction and the damage characteristic frequency band identifier; The feature tensor generation module is used to calculate the thermal stress sensitivity factor based on the strain signal and temperature signal, extract the multi-physics field coupling characteristics associated with the damage characteristic frequency band identifier in the pure signal matrix, and fuse them to generate a high-dimensional damage feature tensor; A risk assessment module is used to input the high-dimensional damage characteristic tensor into a pre-built digital twin risk assessment model. The digital twin risk assessment model locally refines the finite element mesh and updates the material elastic modulus parameters based on the eigenvalues, and outputs an assessment result containing a risk level label and damage location coordinates. The eigenvalues ​​include the spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor. The equipment control module is used to match the control strategy according to the risk level label and generate corresponding equipment control instructions.

[0110] As a further embodiment of the monitoring system, it further includes: The measured data acquisition module is used to obtain the measured vibration frequency domain characteristics at the damage location coordinates; Deviation calculation module, used to calculate the modal confidence factor deviation value between the measured vibration frequency domain characteristics and the simulated stress field in the same spatial coordinates; An over-limit trigger module is used to trigger the constitutive parameter inversion engine when the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times; Iterative optimization module, which is used to iteratively optimize the temperature-elastic modulus mapping table in the material constitutive relationship library based on the covariance matrix adaptive evolutionary algorithm; The iteration log generation module is used to update the optimized temperature-elastic modulus mapping table to the parameter storage area of ​​the digital twin risk assessment model and generate a version iteration log.

[0111] A high-altitude large-scale steel structure corridor safety risk monitoring system in an embodiment of the present application can implement any of the above-mentioned monitoring methods, and the specific working processes of each module in the monitoring system can refer to the corresponding processes in the above-mentioned method embodiments.

[0112] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for example, the division of a module is merely a logical functional division, and in actual implementation, other division methods may be used, such as combining or integrating multiple modules into another system, or ignoring or not implementing certain features.

[0113] The embodiment of the present application also discloses a computer-readable storage medium.

[0114] A computer-readable storage medium stores a computer program that can be loaded by a processor and executed by any one of the above-mentioned methods for monitoring the safety risks of a large-scale high-altitude steel structure corridor.

[0115] Among them, computer-readable storage media can be any tangible medium that contains or stores a program that can be used by or in combination with an instruction execution system, apparatus or device; the program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0116] It should be noted that the computer device and storage medium in the embodiments of this application are, respectively, electronic devices and storage media for use with the aforementioned method for monitoring the safety risks of large, high-altitude steel structure corridors. Therefore, all embodiments of the aforementioned monitoring method are applicable to the computer device and storage medium, and can achieve the same or similar beneficial effects. Since the computer device / storage medium embodiments are generally similar to the method embodiments, their description is relatively simple. For relevant details, please refer to the partial description of the method embodiments.

[0117] In the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0118] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art can understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple situations. A single processor or other unit can implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0119] The above are all preferred embodiments of the present application and are not intended to limit the scope of protection of this application. Unless otherwise specified, any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features. In other words, unless otherwise specified, each feature is merely an example of a series of equivalent or similar features.

Claims

1. A method for monitoring safety risks of large-scale high-altitude steel structure corridors, characterized in that: The monitoring method comprises: The original monitoring data is collected by a multi-source sensor group deployed at each node of the steel structure corridor; the original monitoring data includes strain signals, vibration signals, acoustic emission signals and temperature signals; Performing spatiotemporal alignment processing on the raw monitoring data to generate a time-synchronized sensor data matrix; Dynamically suppress the sensor data matrix based on a pre-trained environmental noise transfer function model, and output a clean signal matrix after noise reduction and a damage characteristic frequency band identifier; Calculating a thermal stress sensitivity factor based on the strain signal and the temperature signal, extracting multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fusing them to generate a high-dimensional damage characteristic tensor; The high-dimensional damage characteristic tensor is input into a pre-built digital twin risk assessment model. The digital twin risk assessment model locally refines the finite element mesh and updates the material elastic modulus parameters based on the characteristic values, and outputs an assessment result including a risk level label and damage location coordinates; wherein the characteristic values ​​include spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor; The control strategy is matched according to the risk level label, and the corresponding device control instructions are generated.

2. A method for monitoring safety risks of a high-altitude large steel structure corridor according to claim 1, characterized in that: The steps of dynamically suppressing the sensor data matrix based on a pre-trained environmental noise transfer function model and outputting a noise-reduced pure signal matrix and a damage characteristic frequency band identifier include: Acquire a time-synchronized sensor data matrix; the sensor data matrix includes vibration signals, acoustic emission signals, wind speed, and temperature and humidity data; Load the pre-trained ambient noise transfer function model; Performing wavelet packet decomposition on the vibration signal and calculating the wavelet packet energy entropy value of each sub-band; According to the energy entropy value, sub-band components with entropy values ​​lower than a set threshold are screened and reconstructed into a primary noise reduction signal; Performing empirical mode decomposition on the primary noise reduction signal to extract intrinsic mode function components; Filtering the intrinsic mode function component related to wind speed based on the environmental noise transfer function model to generate a pure vibration signal; Performing a matching pursuit algorithm on the acoustic emission signal, excluding preset rain noise frequency band atoms when the temperature and humidity data exceeds a preset temperature and humidity threshold, and extracting characteristic atoms that match the crack waveform as crack characteristic waveform atoms; fusing the pure vibration signal with crack characteristic waveform atoms to generate a pure signal matrix; The effective frequency band boundary of the crack characteristic waveform atom is calculated, and a damage characteristic frequency band boundary identifier is output.

3. A method for monitoring safety risks of a high-altitude large steel structure corridor according to claim 2, characterized in that: The steps of establishing the environmental noise transfer function model include: Collect historical background noise data during periods without structural loads; Wind speed data were recorded simultaneously as independent variables; Calculate the wind speed autopower spectrum P xx (f) and wind speed-noise cross power spectrum P xy (f); The frequency domain transfer function is calculated by the cross power spectrum density, and the calculation formula is: H(f)=P xy (f) / P xx (f).

4. A method for monitoring safety risks of a high-altitude large steel structure corridor according to claim 1, characterized in that: The steps of calculating the thermal stress sensitivity factor according to the strain signal and the temperature signal, extracting the multi-physics field coupling feature associated with the damage characteristic frequency band identifier in the pure signal matrix, and fusing them to generate a high-dimensional damage characteristic tensor include: Receive a pure signal matrix and a damage characteristic frequency band identifier; the pure signal matrix includes pure vibration signals and crack characteristic waveform atoms; Synchronously acquire the temporally and spatially aligned raw strain and temperature signals; Based on the strain signal and the temperature signal, solving the partial derivative of strain with respect to temperature within a preset time window to obtain a thermal stress sensitivity factor; Extracting the energy distribution of the crack characteristic waveform atoms in the time domain waveform, and calculating the crack energy entropy value within the damage characteristic frequency band identification range; Performing frequency domain transformation on the pure vibration signal to extract vibration frequency domain features within a preset characteristic frequency band; The thermal stress sensitivity factor, crack energy entropy value and vibration frequency domain characteristics are fused through a 1D-CNN neural network to output a fused high-dimensional damage feature tensor.

5. A method for monitoring safety risks of a high-altitude large steel structure corridor according to any one of claims 1 to 4, characterized in that: The steps of inputting the high-dimensional damage characteristic tensor into a pre-built digital twin risk assessment model, wherein the digital twin risk assessment model locally refines the finite element mesh according to the characteristic value and updates the material elastic modulus parameter, and outputs an assessment result including a risk level label and damage location coordinates include: Receiving the high-dimensional damage feature tensor, including thermal stress sensitivity factor, crack energy entropy value and vibration frequency domain characteristics; Calculating a spatial strain gradient value based on the thermal stress sensitivity factor; When the spatial strain gradient value exceeds a preset threshold, a local refinement operation of the finite element mesh is triggered in the pre-built digital twin risk assessment model; Call the temperature-elastic modulus mapping table in the pre-stored material constitutive relationship library and dynamically update the elastic modulus parameters according to the real-time temperature data; Inputting the crack energy entropy value into a crack growth rate model to calculate a critical damage index; Perform finite element simulation on the refined mesh model after updating the elastic modulus parameters and output the simulated stress field; Based on the preset physical rule threshold, the thermal stress sensitivity factor and the vibration frequency domain characteristics are compared to obtain a first-level decision result; Input the vibration frequency domain features into the pre-trained LSTM time series model, calculate the residual between the predicted value and the actual value, and obtain the secondary decision result; Calculate the modal matching degree between the simulated stress field and the vibration frequency domain characteristics to obtain the three-level decision results; Modifying the preset decision weight ratio according to the critical damage index, and weightedly fusing the first-level decision result, the second-level decision result, and the third-level decision result to generate a risk level label; Extracting peak coordinates in the simulated stress field as damage position coordinates; The risk level label and the damage location coordinates are combined to obtain an assessment result.

6. A method for monitoring safety risks of a high-altitude large steel structure corridor according to claim 5, characterized in that: After the step of outputting the assessment results including the risk level label and the damage location coordinates, the following steps are further included: Extracting the modal confidence factor deviation value between the measured vibration frequency domain characteristics at the damage location coordinates and the simulated stress field at the same spatial coordinates; When the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times, triggering the constitutive parameter inversion engine; Iteratively optimizing a temperature-elastic modulus mapping table in the material constitutive relationship library based on a covariance matrix adaptive evolutionary algorithm; The optimized temperature-elastic modulus mapping table is updated to the parameter storage area of ​​the digital twin risk assessment model, and a version iteration log is generated.

7. A high-altitude large steel structure corridor safety risk monitoring system, characterized in that: The monitoring system comprises: A data acquisition module is used to collect raw monitoring data through a multi-source sensor group deployed at each node of the steel structure corridor; the raw monitoring data includes strain signals, vibration signals, acoustic emission signals and temperature signals; A data processing module is used to perform spatiotemporal alignment processing on the raw monitoring data to generate a time-synchronized sensor data matrix; A noise reduction module is used to dynamically suppress the noise of the sensor data matrix based on a pre-trained environmental noise transfer function model, and output a pure signal matrix after noise reduction and a damage characteristic frequency band identifier; a characteristic tensor generation module, configured to calculate a thermal stress sensitivity factor based on the strain signal and the temperature signal, extract multi-physics field coupling features associated with the damage characteristic frequency band identifier in the pure signal matrix, and fuse them to generate a high-dimensional damage characteristic tensor; a risk assessment module, configured to input the high-dimensional damage characteristic tensor into a pre-built digital twin risk assessment model, wherein the digital twin risk assessment model locally refines the finite element mesh and updates the material elastic modulus parameters based on the characteristic values, and outputs an assessment result including a risk level label and damage location coordinates; wherein the characteristic values ​​include spatial strain gradient values ​​calculated based on the high-dimensional damage characteristic tensor; The device control module is used to match the control strategy according to the risk level label and generate corresponding device control instructions.

8. A high-altitude large steel structure corridor safety risk monitoring system according to claim 7, characterized in that: The monitoring system further comprises: A measured data acquisition module, used to obtain the measured vibration frequency domain characteristics at the damage location coordinates; a deviation calculation module, used to calculate the modal confidence factor deviation value between the measured vibration frequency domain characteristics and the simulated stress field at the same spatial coordinates; An over-limit trigger module is used to trigger the constitutive parameter inversion engine when the modal confidence factor deviation value exceeds the tolerance threshold for a preset number of consecutive times; An iterative optimization module, configured to iteratively optimize a temperature-elastic modulus mapping table in the material constitutive relationship library based on a covariance matrix adaptive evolutionary algorithm; The iteration log generation module is used to update the optimized temperature-elastic modulus mapping table to the parameter storage area of ​​the digital twin risk assessment model and generate a version iteration log.

9. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 6.

Images