Bridge safety early warning method and system based on data model and physical model fusion

Abstract

The invention belongs to the technical field of bridge engineering, and particularly discloses a bridge safety early warning method and system based on fusion of a data model and a physical model. Comprising the following steps: in a first stage, on the basis of a neural network monitoring site excitation-response mapping relation model established by an initial state full-bridge finite element model, calculating response information corresponding to field excitation, and comparing response monitoring signals to determine whether a structure is damaged from a full-bridge angle; in the second stage, through data expansion from a monitoring point position to a non-monitoring point position, a set threshold value is compared to determine a damage position and a damage grade; and in the third stage, adjusting parameters of the model are calculated according to excitation-response signals of a monitoring point position sensor, the finite element structure model is updated, the residual bearing capacity of the structure is calculated by combining state information borne by the structure model, and the service life of the component is determined. The invention provides a safety assessment method fusing monitoring data and a physical model for structural health monitoring, and the method is clear in physical significance, high in real-time performance and reliable in result.

Description

Technical Field

[0001] The present invention belongs to the technical field of bridge engineering, and more specifically, relates to a bridge safety early warning method and system based on the fusion of data model and physical model. Background Art

[0002] With the opening of cross-sea cluster projects such as the Hong Kong-Zhuhai-Macao Bridge in recent years, "factory-based" and "standardized" prefabricated assembled structures have been increasingly widely used in bridge construction. As a result, the question is how to achieve safety assessment and early warning for bridge structures with continuous spans of tens of kilometers through a set of "standardized" and "intelligent" health monitoring systems. In the safety assessment of bridge model correction structures, there are mainly the following two issues: First, during the service of prestressed concrete bridges, the offshore environment has a prominent effect and the load type is highly nonlinear. In the offshore environment, reinforced concrete is prone to chloride ion corrosion, which changes the pH of the concrete body. Under this influence, the prestress relaxation loss of offshore bridge structures is aggravated compared to inland areas. For load types, the cables are subjected to wind-induced vibrations, the temperature difference produces thermal expansion and contraction, and the movement of vehicles causes dynamic evolution of the structure. Such complex environmental factors and load conditions will lead to a decrease in the real-time performance of structural calculations.

[0003] Secondly, the physical model information obtained by calculating the structural components in different parts is diverse, and the heterogeneity of the information affects the calculation speed of the structural safety assessment and the real-time generation of the assessment results. For the diversity of physical model information, finite element models of different scales should be used for targeted analysis, and corresponding safety assessment methods should be proposed to evaluate it.

[0004] Therefore, in view of the causes of performance degradation and the upper loads on offshore prestressed concrete bridges, a complete on-site non-destructive monitoring system is designed in a standardized manner. The damage information of the unmonitored points of the structure can be inferred based on the information collected by the non-destructive monitoring system at the on-site monitoring points. Mining the potential loads of the structure is not only a development trend of intelligent bridge monitoring and management, but also a rigid demand. Due to the significant degree of "factoryization" and "standardization" of offshore prestressed concrete bridge structures, the "standardization" of the monitoring process is also very critical. This helps to increase the speed of mapping data to physical models, establish a bridge operation and maintenance model that integrates data and physics, evaluate the service status of the structure, and put forward reliable maintenance opinions and suggestions.

[0005] In recent years, with the development of science and technology, many monitoring methods and evaluation systems have developed in the direction of improving the "intelligence" of the bridge engineering maintenance process. In the field of digital-physical integration, the technical development in model correction is as follows: At present, model correction tends to use computer-aided means to establish the mapping relationship between sensor signals (excitation, response information) and model information (stiffness matrix, structural vibration mode), thereby correcting the finite element model to analyze the structural state or to directly quantify the calculation of structural safety and conduct state assessment. At this stage, the main analysis method is to establish the mapping relationship between signals and model information through PSO, BP and other means. The analysis speed is fast, but the physical and mechanical information between input and output is unclear during the analysis process, and it is difficult for the network to respond to erroneous information in a timely manner. In the field of safety assessment, the hierarchical analysis method and expert evaluation method are mainly used to evaluate the safety status of bridge structures, assisted by big data analysis methods such as neural networks. This method can effectively achieve rapid evaluation through cloud computing and server analysis methods, but the evaluation method lacks dynamic evolution capabilities. Most importantly, in the process of safety assessment, a single type of load is usually evaluated, and the situation of dynamic coupling of multiple load types through physical relationships is not considered, which lacks accuracy in actual engineering applications.

[0006] However, the current assessment process is difficult to achieve real-time processing of on-site physical model structural information, analyze the actual stress state of the structure under the current state, evaluate the degradation trend of the material, and calculate the residual bearing capacity and stress redistribution of the structure. A fully functional offshore prestressed concrete bridge structure health monitoring system is urgently needed to achieve real-time structural correction and safety status assessment in multi-source data fusion. Summary of the invention

[0007] In view of the above defects or improvement needs of the prior art, the present invention provides a bridge safety early warning method and system based on the fusion of data model and physical model, which realizes real-time analysis of sensor monitoring information through proxy model, processes and mines structural model information, evaluates structural safety status, and maps the proxy model calculation results to the built finite element model, so as to realize real-time correction of the finite element model, display the current state of the structure, and improve the accuracy of structural performance analysis. Through finite element analysis, not only can the location and damage level of potential damage be discovered, but also various types of loads can be applied to the finite element model in advance, the neural network analysis data set can be expanded, and a reliable real-time updated training set can be provided for neural network pre-training, forming a real-time excitation-load mapping relationship with perfect physical relationship, and improving the accuracy of neural network output data.

[0008] To achieve the above object, according to one aspect of the present invention, a bridge safety early warning method based on the fusion of data model and physical model is proposed, comprising the following steps: Step S1, constructing a multi-scale finite element model of the bridge structure, calculating the structural response by inputting simulated excitation data into the finite element model, and generating a data set including an excitation signal, a response signal and structural model information; Step S2, constructing a neural network model, and using corresponding data sets to train the neural network model according to different neural network calculation purposes, the neural network model includes: neural network FR, neural network RR and neural network FR-FE; Step S3, first safety assessment stage: using the neural network FR to calculate the on-site monitoring point excitation-monitoring point response, and obtain the calculated response based on the initial state finite element model to determine whether the bridge structure is abnormal or damaged; Step S4, second safety assessment stage: using neural network RR to analyze the monitoring point response information input in step S3, expanding the non-monitoring point calculation response, and performing calculation response monitoring threshold level classification to determine the location and damage level of the component defined as damaged; Step S5, the third safety assessment stage: the excitation signal-response signal collected on site is analyzed by the neural network FR-FE to obtain the correction parameters for the updated finite element model, and the overall bearing capacity level of the bridge and the remaining service life of the structure are calculated based on the updated multi-scale finite element model.

[0009] As further preferred, step S1 includes the following steps: Step S11, dividing the construction section where a refined model needs to be established according to the most unfavorable position of the bridge and the position of the pre-buried sensors on site; Step S12, dividing the finite element structure into multiple components according to the sensor data type and layout points, establishing component scale and full bridge scale finite element models according to the characteristic information provided by the engineering drawings, using different grid scales in the same finite element model to simulate monitoring components and non-monitoring components, and establishing a multi-scale finite element model; Step S13, using a multi-scale finite element model to simulate and analyze the main loads and performance changes of the bridge, and establishing a data set by exporting the calculation information and excitation information of different positions of the initial structural finite element model.

[0010] As a further preferred embodiment, in step S13, the structural response is calculated by inputting simulated excitation data into the finite element model excitation monitoring point, that is, a suitable degradation model is selected for different structural models, a bridge geometric parameter degradation model is established, and the degradation range of the finite element model calculation parameters compared with the initial model is determined.

[0011] As further preferred, in step S2, by inputting simulated loads to the excitation monitoring points in the full-bridge scale finite element model in the initial state, the calculated data of the response monitoring points are extracted as a training data set for the neural network FR; If the bridge structure is judged to be intact, the non-monitoring point response signals of the neural network RR data set are obtained by inputting the simulated load into the full-bridge scale finite element model in the initial state; if the bridge structure is judged to be damaged, its non-monitoring point response signals are obtained by inputting the simulated load into the revised full-bridge scale finite element model, and according to the assessment status of the first safety assessment stage, the corresponding node response data are extracted to obtain the monitoring point response-unmonitoring point response training set; Based on the model of the multi-scale finite element model, the excitation sensor data and response sensor data in each finite element model node are integrated to obtain the FR-FE training set.

[0012] As further preferred, step S3 includes the following steps: Step S31, dividing the signals collected on-site from the excitation-response signals of the on-site monitoring point sensors into two parts, one part of the monitoring point excitation is used as a training process of the neural network FR, and the other part of the monitoring point response is used as a verification set with the neural network FR; Step S32, inputting the excitation monitoring point signal in step S31 into the full-bridge finite element model to obtain a neural network FR, and obtaining a response monitoring point calculation response signal based on the initial finite element model; Step S33, comparing the calculated response signal of the response monitoring point in step S32 with the on-site monitoring response signal in step S31, calculating the mean square error between the measured data and the calculated data of all the response monitoring points, and determining whether the signal obtained by calculating the initial state full-bridge scale finite element model of the verification signal is consistent with the on-site monitoring point signal; Step S34: If the signal comparison error is within the given error range, it means that the offshore bridge structure has not been damaged at a macroscopic level, and the macroscopic safety status of the structure is good; if the signal comparison error exceeds the given range, it means that the structure can be judged as damaged at a macroscopic level through the calculation of the field sensor signal by the neural network FR; Step S35 uses the neural network FR to calculate the on-site monitoring point excitation-monitoring point response, and obtains the calculated response based on the initial state finite element model to determine whether the bridge structure is abnormal or damaged.

[0013] As further preferred, step S4 includes the following steps: Step S41, calculate the current structural model state assessment parameter according to the calculation response of all structural nodes of the full bridge scale obtained by the neural network RR and the set warning threshold , and evaluate the state parameters The damage level is divided according to the value of The threshold value of Step S42: Arrange the structural damage levels in step S41 according to their magnitude, and analyze the state assessment parameters of the most dangerous position of the structure. Whether it exceeds the given warning threshold range; Step S43: If the status evaluation parameter If the warning threshold is exceeded, the damage position and its corresponding damage level need to be located; otherwise, if all node status evaluation parameters If the warning threshold is not exceeded, all nodes in the entire structure are in a safe state.

[0014] As a further preferred embodiment, in step S41, the state evaluation parameter Divided into: <2.0%, 2.0%< <5.0%, 5.0%< <20.0%, >20.0% four levels.

[0015] Step S5 includes the following steps: Step S51, input the monitoring point excitation-monitoring point response signal of the field sensor into the neural network FR-FE, and extract the weight matrix of the hidden layer of the neural network; Step S52, based on the inversion function in step S2, calculating the calculation parameters of the multi-scale finite element model, correcting the finite element model in real time, using the corrected multi-scale finite element model to calculate the residual bearing capacity of the structure, and evaluating the overall safety information of the structure; Step S53, calculating the material degradation index at each node based on the modified vector finite element model parameters and the initial vector finite element model parameters , structural cross-sectional area degradation index and the degradation index of the moment of inertia of the structural section According to the on-site environmental factors data obtained by on-site sensor monitoring, the degradation model of material properties in the degradation index is derived, and the material degradation index is used to determine the degradation model of the material properties in the degradation index. Predict the remaining useful life of materials, fit , With the time-varying curve, the correlation between the reduction of the cross-sectional properties and time is derived to obtain the degradation function of the cross-sectional properties; Step S54, evaluating the refined model part of the finite element model according to the material degradation index , structural cross-sectional area degradation index and the degradation index of the moment of inertia of the structural section The finite element component scale model is modified, and the component safety information is evaluated and the residual bearing capacity of the component is calculated based on the finite element model calculation results.

[0016] As further preferred, the following steps are also included: The finite element model is directly corrected using the structural damage information directly monitored by the on-site sensors in S3. That is, the final correction of the finite element model needs to be based on the consistency between the structural damage and the on-site monitoring damage. The finite element model is updated according to the degradation coefficients of the section and material analyzed by the monitoring signal in S5 and the adjustment parameters of the corrected structural damage. The updated finite element structural model is used to generate a new data set for training the neural network model and correcting the finite element model.

[0017] According to another aspect of the present invention, a bridge safety early warning system based on the fusion of a data model and a physical model is provided, comprising: The first main control module is used to construct a multi-scale finite element model of the bridge structure, calculate the structural response by inputting simulated excitation data into the finite element model, and generate a data set; A second main control module is used to construct a neural network model and use the data set to train the neural network model, wherein the neural network model includes: a neural network FR, a neural network RR, and a neural network FR-FE; The third main control module is used to calculate the on-site monitoring point excitation-monitoring point response using the neural network FR, obtain the calculated response based on the initial state finite element model, and conduct comparative analysis to determine whether the bridge structure is abnormal or damaged; The fourth main control module is used to use the neural network RR to analyze the response information of the monitoring points input in the third main control module, expand the calculated responses of the non-monitoring points, and classify them to determine the damage level and the location of the damage parameter exceeding the predetermined threshold; The fifth main control module is used to analyze the excitation signal-response signal collected on site through the neural network FR-FE, obtain the correction parameters for updating the finite element model, and calculate the overall bearing capacity level of the bridge and the remaining service life of the structure based on the updated multi-scale finite element model.

[0018] In general, the above technical solution conceived by the present invention has the following technical advantages compared with the prior art: 1. The three-stage bridge safety early warning method based on the fusion of data model and physical model can quickly and comprehensively analyze the on-site sensor signals, give early warning of the threshold of the structural health status through the excitation signal and response signal, and modify the finite element model to comprehensively and carefully evaluate the safety performance and service life of the entire structure / components, thereby improving the efficiency of risk assessment and ensuring the accuracy of the assessment results.

[0019] 2. The present invention can realize the real-time dynamic update of the structural model on the finite element calculation platform or the structural information management platform, so that the supervisor can obtain information to judge the structural status during the periodic structural safety inspection. Similarly, after the periodic structural safety inspection, the test report data can be compared to verify the data reliability of both the platform model and the report content.

[0020] 3. The present invention generates a training set by updating the finite element model in real time to establish a neural network model for rapid processing of field signals, realizes rapid processing of field signals based on different purposes, and evaluates the structure in three stages. Each evaluation stage targets different safety assessment conditions of the structure, and by setting early warning thresholds, the damage and potential damage detected can be fed back in a timely and quantitative manner.

[0021] 4. The safety assessment process proposed in the present invention is to calculate the remaining service life of the structural components based on the statistical prediction value of the load excitation monitored by the sensor; and to evaluate the remaining bearing capacity of the structure based on the modified finite element model. The finite element model serves as a reliable physical and mechanical basis, and the multi-scale simulation displays the damage information of the structure through the finite element model of the component scale. The structural state variables (stiffness matrix, structural vibration mode) revealed by the finite element model are used to evaluate the safety status of the bridge as a whole and its components. The physical meaning of the evaluation process is clear and the evaluation results are reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention is a flow chart of a bridge safety early warning method based on the fusion of a data model and a physical model involved in an embodiment of the present invention; Figure 2 It is a schematic diagram of a multi-scale finite element model established by taking a prestressed simply supported beam as an example according to an embodiment of the present invention; Figure 3 This is a flow chart of the first stage security assessment involved in an embodiment of the present invention; Figure 4 This is a flow chart of the second stage security assessment involved in an embodiment of the present invention; Figure 5 This is a flow chart of the third stage security assessment involved in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0024] Example 1 like Figures 1 to 5As shown, the present embodiment provides a bridge safety early warning method based on the fusion of data model and physical model, which adopts a three-stage safety assessment method for structural assessment. In the first stage, the neural network monitoring point excitation-response mapping relationship model is established based on the initial state full-bridge finite element model, and the response information obtained by the on-site sensor excitation is quickly calculated. The monitoring signal is compared to determine whether the structure is damaged from the perspective of the whole bridge. After the damage is determined in the first stage, the second stage is entered, and the damage location and damage level are determined by comparing the set threshold through data expansion from the monitoring point to the non-monitoring point. Regardless of whether the second stage is entered, the third stage will be entered. First, according to the adjustment parameters of the monitoring point sensor excitation-response signal calculation model, the finite element structure model is updated, the residual bearing capacity of the structure is calculated by the numerical calculation method of calculating the structural membership, and the service life of the component is determined by the uncertainty method of probability statistics.

[0025] This embodiment intends to use neural network as a technical means, replacing the healthy initial finite element model in the first stage of the evaluation, and establishing a mapping relationship between the structural excitation of the monitoring point and the structural response of the monitoring point; in the second stage, based on the updated finite element model, the neural network is used to calculate and derive the non-monitoring point data through the monitoring point data, and the data position of the non-monitoring point is expanded; the third stage is used to analyze the excitation-response signal of the field sensor, and calculate the calculation parameters of the finite element model through the inversion function. And according to the damage directly monitored by the field sensor, the full bridge and component scale finite element model is comprehensively updated.

[0026] This embodiment intends to use a neural network to generate monitoring excitation data to model response information, which can improve the speed of analyzing the monitoring excitation information of the field sensor. A neural network is used to establish a mapping relationship. In the first stage, the initial state multi-scale finite element model is used to calculate the external excitation. In the second stage, the updated finite element model is used to establish a mapping relationship between the monitoring point and the unmonitored point information. In the third stage, the neural network is used to correct the finite element model by establishing a data-physical model that reveals the relationship between the "excitation signal-response signal" and the "finite element model parameters". The use of a neural network can improve the speed of analyzing the monitoring signal.

[0027] The safety assessment method of the multi-scale finite element model of the bridge structure provided in this embodiment can realize real-time analysis of sensor monitoring information through the proxy model, process and mine the structural model information, evaluate the structural safety status, and feedback the calculation results of the proxy model to the built finite element model to realize real-time correction of the finite element model, display the current state of the structure, and improve the accuracy of structural performance analysis. Through finite element analysis, not only can the location and damage level of potential damage be discovered, but also various types of loads can be applied to the finite element model in advance, the neural network analysis data set can be expanded, and a reliable real-time updated training set can be provided for neural network pre-training, forming a real-time excitation-load mapping relationship with perfect physical relationship, and improving the accuracy of neural network output data.

[0028] In addition, the present embodiment relates to a structural safety assessment system, including: a multi-scale finite element model established at any time, a deep learning-based proxy model established at any time, a neural network of on-site monitoring point excitation-monitoring point response mapping relationship, a neural network of on-site monitoring point response-unmonitored point response mapping relationship, and a loss function of the proxy model calculated based on proxy model analysis information and on-site monitoring information.

[0029] In any of the above embodiments, the structural monitoring excitation information is mainly vehicle load, temperature load, wind load and earthquake action. The structural monitoring response information is mainly the structural stress, strain and overall structural deflection (prestress-related and other monitoring information) of the bridge monitoring point obtained by other non-destructive testing methods or pre-embedded means. When establishing the offshore bridge structure model, the multi-scale finite element module establishes structural models of different scales according to the force characteristics of different sensor detection parts and structures, and establishes the mapping relationship between the finite element structural model and the sensor monitoring signal through a numerical method, including: the full-bridge scale finite element model describes the overall behavior and force mode of the bridge structure; based on the established full-bridge scale model, the stress distribution and deformation of the structure as a whole are calculated, the main purpose of which is to analyze the mechanical parameter distribution (stress, strain, deformation state, etc.) and structural behavior (deflection, collapse, etc.) of the structure under the action of external loads, as the calculation basis of the neural network numerical model; the refined finite element structural model represents the microscopic description and simulation of the damage on the structural components, establishes the connection between the calculation parameters and the full-bridge scale finite element model, and assists in correcting the full-bridge scale finite element model.

[0030] Based on the signals monitored on-site by sensors, the data of unmonitored points are expanded and analyzed through monitoring points, and threshold warnings are set through static indicators (such as displacement, stress, strain, etc.).

[0031] The finite element model is corrected in real time through the neural network, and the safety status of the physical structure is evaluated based on the finite element model, including: the mechanical evaluation index of the structure, mainly the bearing capacity of the multi-scale finite element model, and the direction is mainly bending, shear, tensile bearing capacity, etc. The residual bearing capacity of the structure is mainly derived from the remaining service life of the material based on the material degradation model, and the degradation function of the section is derived based on the fitted section degradation parameters.

[0032] Based on a combination of any of the above solutions, this embodiment is implemented through the following technical solutions: S1: Establish a multi-scale finite element model to correct the bridge model at the initial moment, and make sure that the model information is consistent with the actual engineering structure. In terms of modeling scale, the unit is divided into parts that need to be divided into macro-scale structures and refined structures according to the layout of the monitoring points. Different grid scales are used in the same finite element model to simulate monitoring components and non-monitoring components, and a multi-scale finite element model is established. The structural response is calculated by inputting simulated excitation data into the finite element model excitation monitoring points, and a data set for subsequent neural network training is pre-generated to provide model support for subsequent neural network training.

[0033] S2: Establish a neural network and use a multi-scale finite element model to generate a data set for pre-training. The established multi-scale finite element model is used to generate data containing [excitation monitoring point signal-all node response signals-finite element model parameters] by inputting simulated excitation data for pre-training of the neural network required for each evaluation stage. The neural network FR quickly calculates the relationship between the excitation monitoring point signal-monitoring point response signal, and the neural network RR expands the monitoring point response signal-non-monitoring point response signal through real-time signals. The neural network FR-FE is used to analyze the [excitation signal-response signal] and [finite element model parameters] information. The calculation process of the model needs to meet the actual bridge experiment or scaled test to verify the applicability of the neural network.

[0034] S3: Achieve macro early warning by calculating the excitation monitoring point signal-response monitoring point signal through neural network FR. Use neural network FR to calculate the excitation monitoring point-response monitoring point signal and obtain the calculated response based on the initial state finite element model. Compare the calculated response of the structure with the response information of the on-site monitoring point to see if there is a big difference, and determine whether the bridge structure is abnormal or damaged.

[0035] S4: The non-monitoring response signal is expanded from the response monitoring point signal through the neural network RR to realize the damage warning of the most unfavorable position of the component. According to the calculation results of the initial refined finite element model and the design load, the non-monitoring point response signal value is divided into multiple levels. The calculated non-monitoring point response signal is used to evaluate the damage level and locate the highest damage level (the most dangerous position).

[0036] S5: Use the neural network FR-FE data-physical model to analyze the relationship between [excitation signal-response signal] and [finite element model parameters], modify the multi-scale finite element model in real time, and evaluate the safety status of the structure / component. Structural / component bearing capacity calculation and safety assessment: In this assessment stage, the residual bearing capacity of the structure will be calculated from the overall structure and the refined components, the remaining service life will be analyzed, and the safety status of the structure will be evaluated. First, the neural network FR-FE analyzes the [excitation signal-response signal] collected on site to obtain the [finite element model parameters] to update the finite element model. Use the multi-scale finite element model to calculate the overall bearing capacity level; isolate the refined component finite element model and evaluate the remaining service life.

[0037] The method further comprises the following steps: S6: Use the structural damage information directly obtained by the sensor to directly correct the finite element model. Use the on-site sensor to monitor the structural damage information to directly correct the multi-scale finite element model. The final correction result of the finite element model in this iteration step includes the direct damage read by the sensor and the correction result based on the data analysis of the excitation monitoring point and the response monitoring point. Use the finite element model parameters in S5 to remove the influence of the directly corrected damaged parts on the parameters, and update the final corrected finite element model of this iteration step. Use the updated finite element structural model to generate a data set for subsequent neural network pre-training and finite element model correction.

[0038] As a further improvement of the above solution, S1 specifically includes: 1) Establish a finite element model based on the engineering design drawings, and select appropriate unit types, unit lengths, and boundary conditions to establish finite element models of the entire bridge scale and component scale.

[0039] 2) Adjust the transition part of the two meshing scales of the finite element model. If appropriate, the selected mesh scale and simulation method can be verified by load testing, and the finite element model can be adjusted so that it can output the calculated response of the response monitoring point position.

[0040] 3) Use multi-scale finite element models to simulate and analyze the main loads and performance changes of the bridge (vehicle loads, temperature changes, wind loads, prestress loss, concrete performance degradation, etc.), and establish a neural network analysis data set by exporting the calculation information and excitation information of different positions of the initial structural finite element model. The structural monitoring excitation information mainly includes vehicle loads, temperature loads, wind loads and earthquake effects. The structural monitoring response information mainly includes the structural stress, strain, and overall structural deflection (prestress-related and other monitoring information) of the bridge monitoring point obtained through other non-destructive testing methods or pre-embedded means.

[0041] As a further supplement to the above scheme, as a significant environmental factor for offshore bridges compared to other bridges, the chloride ion monitoring density will directly act on the model information correction through the material degradation model, etc., so it is not included as the main load in the main assessment process.

[0042] As a further supplement to the above scheme, in order to obtain training sets for different neural networks, it is necessary to establish data sets through different finite element models. For example, in the neural network FR, it is necessary to obtain the monitoring point excitation-monitoring point response of the initial full-bridge scale finite element model; in the neural network FR, it is necessary to obtain the monitoring point response-non-monitoring point response of the refined partial finite element model.

[0043] Furthermore, the S2 specifically includes: 1) Divide the data set in S1 into two parts, one as the training set of the neural network and the other as the training set of the neural network. Flatten the initial data set to adapt it to the sensor as a suitable signal input on the time scale. If it is for the neural network FR, only the monitoring point excitation-response signal is required. If it is for the neural network RR, the finite element model of the previous time step needs to be used to calculate the non-monitoring point response based on the monitoring point response.

[0044] 2) Select a suitable neural network method, analyze the data divided into training set and validation set in 1), adjust the loss function according to the output data type, and obtain the initial pre-trained neural network model.

[0045] 3) In this stage, a total of three neural networks need to be established. The neural network FR establishes the mapping relationship between the excitation monitoring point signal and the response monitoring point signal based on the full-bridge scale model; the neural network RR establishes the response monitoring point-non-monitoring point response mapping based on the finite element model of the component part; the neural network FR-FE is used to analyze the excitation monitoring point signal-response monitoring point signal, calculate the calculation parameters of the finite element model, and use it for finite element model correction. The capital letters in the formed neural network FR-FE are the abbreviations of force and response, and the latter FE represents the finite element model.

[0046] 4) Analyze the physical meaning of the weight matrix values ​​in the neural network FR-FE, and propose a calculation function for weight inversion of structural model information. Improve the physical information inversion part of the neural network to verify the credibility of the neural network FR-FE.

[0047] 5) Use the validation set divided in 1) to verify the neural network and check the accuracy of the inversion function calculation in 4). If the error of the inversion function inferring the structural model information does not meet the preset range, return to 4) to re-derive the inversion function of the physical information. Use scaled model tests or real bridge tests to verify the applicability of the inversion function and adjust the parameters.

[0048] As a further supplement to the above solution, model interpretation methods such as SHAP, LIME, and feature attribution can be used on neural networks to improve the interpretability of the network and enhance the physical meaning of the code.

[0049] As a further supplement to the above scheme, if damage is determined at the scale of the entire bridge in several consecutive time steps, it can be determined that the structure is abnormal or damaged at the scale of the entire bridge. In the next time step, the assessment phase I can be skipped directly, and the assessment phase II can be entered to locate the damaged part of the component scale model and issue a threshold warning.

[0050] Furthermore, the S4 specifically includes: 1) Since the conclusion obtained in S3 is that the bridge structure has been damaged, the finite element component scale model can directly obtain the damage state of the structure based on the non-destructive monitoring device arranged on site and make preliminary corrections to the component scale model.

[0051] 2) Input the data of the monitoring point of the calculated on-site sensor response into the neural network RR, and calculate the calculated response data of the non-monitoring point. Expand the response signal from the monitoring point to the non-monitoring point.

[0052] 3) Extract the monitoring response point data from the node calculation response data, compare the calculation data with the sensor monitoring data, and calculate the current model response evaluation parameters The physical meaning of the calculation parameter correction data is , The finite element model calculates the stress, strain, deflection and other responses of the bridge when it is grounded and opened to traffic at the initial moment. Indicates the stress, strain, deflection and other response information of the current sensor signal. State assessment parameters Divide into <2.0%, 2.0%< <5.0%, 5.0%< <20.0%, >20.0% four levels.

[0053] As a further supplement to the above scheme, the component scale model state assessment parameters Calculation, denominator The corresponding structure is the corrected structural model of the previous time step.

[0054] Optional, The calculation can recalculate the initial response of the structure based on a modified structural model based on adjustments made during construction.

[0055] Furthermore, the S5 specifically includes: 1) Input the monitoring point excitation-monitoring point response signal of the field sensor into the neural network FR-FE and extract the weight matrix of the hidden layer of the neural network.

[0056] 2) The material degradation index of each node of the modified vector finite element model parameters and the original vector finite element model parameters calculated based on the inversion function in S2 , structural cross-sectional area degradation index and the degradation index of the moment of inertia of the structural section , the modified multi-scale finite element model is used to calculate the residual bearing capacity of the structure and evaluate the overall safety information of the structure.

[0057] 3) Based on the on-site environmental factors data obtained by on-site sensor monitoring, derive the material degradation model and analyze The significance of this model is to predict its remaining useful life. , With the time-varying curve, the correlation between the reduction of cross-sectional properties and time is derived, and its remaining service life is predicted.

[0058] 4) For the modification of the refined model, the model is first modified directly through the damage information obtained by on-site sensor monitoring, and then the degradation index is used to modify the model. , , The finite element component scale model is modified, and the component safety information is evaluated and the residual bearing capacity of the component is calculated based on the finite element model calculation results.

[0059] 5) Since the refined finite element model in S4 is directly corrected by the nondestructive testing results of the on-site sensors, the refined finite element model can be used to obtain the quantitative information of defects (concrete crack width, length, depth and location, prestressed steel bar section reduction and corrosion area) and obtain a quantitative damage assessment report.

[0060] 6) The modified component scale model is evaluated by the state parameters , the degree of degradation of its resistance is quantitatively evaluated based on the degradation model and the fitted degradation curve.

[0061] Furthermore, the S6 specifically includes: 1) According to the modified finite element full-bridge scale model and component scale model described in S5, a new data set is generated according to the load type and performance change (vehicle load, temperature change, prestress loss, concrete performance degradation, etc.) in S1 to participate in the generation and iteration of neural network FR, neural network RR and neural network FR-FE; 2) The updated neural network FR-FE not only needs to meet the pre-set loss function requirements, but also needs to add some finite element model information (such as boundary conditions or finite element unit types) as the loss function. If the ratio of the finite element model calculation parameters calculated by S2 to the calculation parameters of the corresponding position of the initial model is greater than 1, it means that the material performance at this position is better than the initial state. Then this parameter does not conform to physical common sense and needs to be corrected.

[0062] Furthermore, the model building and correction process is calculated through the cloud platform, and the collected sensor data will also be synchronized to the cloud server.

[0063] Furthermore, the evaluation process ultimately results in a cloud document report on the structural status, in which the finite element model is directly corrected with concrete crack information and prestressed tendon defects, the excitation part is mainly vehicle load, and the response part includes but is not limited to displacement, inclination and strain collected by on-site sensors, and the structural natural frequency and modal information obtained based on the synchronously corrected finite element model, and the evaluation includes cable force and vehicle load.

[0064] Furthermore, the monitoring points of the bridge refer to: according to the actual structure type, by applying the measured vehicle load through the finite element model, etc., selecting the key points with large deflection and stress of the whole bridge, or the points that can reflect the dynamic characteristics.

[0065] Example 2 The following describes in detail a specific implementation example of the present invention in conjunction with specific drawings, and explains in detail a safety assessment method for a bridge model correction structure of the present invention.

[0066] It should be noted that the finite element model used can be established by commercial finite element software Midas, Abaqus, ANSYS, etc., or by MATLAB. The use of finite element software is not limited here, and the finite element in the present invention only provides the mechanical calculation results of the load-excitation of the offshore bridge structure, which is used to establish the neural network data set and the third-stage safety assessment. In addition, the neural network is used in the present invention to replace the finite element model to directly obtain the calculation error between the calculated response and the measured response through the sensor excitation signal and the model calculation parameters. The neural network can use CNN, RNN, GAN and other branches. Each neural network architecture has different characteristics during use, and this example does not make specific restrictions on this. The on-site monitoring load information obtained by this method does not involve the personal information of the owner, and uploading the information to the cloud will not infringe on personal privacy. The collected and uploaded monitoring information only includes the vehicle axle weight and the driving position and driving speed of the vehicle.

[0067] It should be noted that the method provided by the present invention is not limited to newly built offshore prestressed concrete bridges, but is also applicable to existing bridges whose regular monitoring reports can provide complete information on bridge structure damage. For existing bridges, it is impossible to obtain model information before they were built and opened to traffic, so the structural model used as a reference for structural correction comparison is the structural model on the design drawing. Before updating the model at the next time step, more detailed multiple iterations are required to update the initial finite element model so that it is updated and modified to meet the current working status.

[0068] like Figure 1 As shown, the present invention provides a three-stage bridge safety early warning method based on the fusion of data model and physical model. The following case is the evaluation process of a prestressed concrete continuous rigid frame bridge by this evaluation method, which specifically includes the following steps: S1: The main load-bearing components of prestressed concrete are reinforced concrete beams, prestressed steel bars and piers connected to the superstructure. The beams and piers are divided into different construction sections according to the cross-sectional size. According to the most unfavorable position calculated by the finite element model and the position of the embedded sensors on site, the construction sections that need to establish a refined model are divided. The finite element structure is divided into multiple components according to the sensor data type and layout point, and a multi-scale finite element model is established according to the characteristic information provided by the engineering drawings. The full-bridge scale part model in the multi-scale reveals the maximum bending moment, shear force and deformation information of the cross section within the entire bridge range. The component scale part deeply calculates the stress and strain distribution of the cross section, as well as the relevant mechanical model of the prestressed tendons as the main load-bearing components and the surrounding concrete. In addition to the finite element model, a model reflecting the component scale to the full-bridge scale needs to be established. The reaction relationship here is mainly the conversion of finite element model information at different modeling scales.

[0069] S2: Pre-train neural networks to establish proxy models for different physical relationships. The neural network FR is used to provide the physical relationship between the excitation monitoring point and the response monitoring point, the neural network RR is used to provide the physical relationship between the response monitoring point and the response of the non-monitored point, and the neural network FR-FE is used to analyze the sensor excitation monitoring point-response monitoring point signal and directly generate the calculation parameters of the finite element model. The generalization process of the monitoring point response-non-monitoring point response can be calculated by the finite element model established in S1.

[0070] S3: Safety assessment phase I: Use the neural network FR to calculate the monitoring point input. At this time, the input data of the neural network FR is all the excitation signals of the field sensors, and obtain the calculated response of the set response monitoring point position based on the model information of the initial state finite element model. Calculate the overall mean square error MSE between the output response and the measured response data of the response monitoring point, and analyze whether the physical information of the structure has changed with the physical information in the finite element main model in the previous time step through the response signal, so as to determine whether the bridge structure is abnormal or damaged.

[0071] S4: Safety assessment phase II: If the bridge structure is judged to be not intact in S3, the assessment phase enters phase II. In this phase, the neural network RR is used to calculate the input response monitoring point signal in S3 to obtain the calculated response of the non-monitoring point. Since the physical information has a correlation under the constraints of the finite element model, the structural response assessment parameters are divided into four levels according to the physical information of the initial refined finite element model. The calculated assessment parameters are classified into a certain level and the damage is located, and the node safety assessment status parameters and warning positions are sorted.

[0072] S5: Safety Assessment Phase III: Regardless of whether the assessment phase enters Assessment Phase II, it will enter Assessment Phase III. In this assessment phase, the residual bearing capacity of the structure will be quantitatively and qualitatively analyzed from the perspective of the entire bridge and components. First, the finite element model is corrected by the neural network FR-FE. For the refined component part of the finite element model, the residual bearing capacity quantitatively evaluated by the finite element model is used to calculate the remaining service life of the structure based on the material degradation model proposed in S1; for the finite element model of the entire bridge, the overall bearing capacity level of the structure is directly calculated to quantitatively evaluate the safety status of the model parameter correction coefficient, fit the geometric property degradation model, and statistically predict the load exceedance probability using the load information collected on site.

[0073] S6: Use the three-dimensional characteristics of the cracks in the concrete structure in the on-site monitoring information to modify the concrete beam model, use the prestressed tendon monitoring information to modify the elastic modulus and tension of the prestressed tendons, and directly modify the finite element refined structural model. And adjust the calculation parameters of the full bridge model through the modified mesoscale model. Divide the degradation coefficient of the model information analysis section and material obtained by back calculation in S5 by the directly modified equivalent structure adjustment parameter, update the finite element model of this time step, and use the updated finite element structural model for subsequent neural network pre-training and finite element model modification.

[0074] Step S1 includes three steps: S11 to S13.

[0075] S11: A multi-scale finite element model was established based on the structural construction drawings. The beam element was selected as the main element of the full-bridge scale model. The solid element was used to establish a refined model of the concrete. The rod element was used to simulate the prestressed steel bars in the concrete. A bond-slip model was added to simulate the interaction between the two.

[0076] S12: Analyze the bridge model and select appropriate degradation models for different structural models. For the concrete part, steel part and prestressed tendon part, different degradation models are selected based on the material properties. Choose an appropriate method to establish the bridge geometric parameter degradation model, and further determine the degradation range of the finite element model calculation parameters compared with the initial model when each iterative step in the proxy model in S4 is updated.

[0077] S13: Define the response signal warning threshold of each node in the threshold evaluation in S4, and define the component remaining service life prediction model through the statistical load information distribution method in S5.

[0078] For step S13, since the structure monitored in the design case is a continuous steel structure, the key points are set to the intersection positions of 1 / 4, 1 / 2, and 3 / 4 of the beam length and the beam pier. Furthermore, the response monitoring points need to be encrypted at 1 / 4L of the side span and in the middle of the main span to improve the correction accuracy; if the bridge structure form of the design monitoring is a simply supported one, the key points set in the influencing matrix are 1 / 4, 1 / 2, and 3 / 4 of the beam length and the position of the support, and the key monitoring positions of strain, inclination, and deflection will be focused on the key points; if the bridge structure form is continuous, in addition to arranging the key points on the main span, it is also necessary to add key points at the side span position.

[0079] like Figure 2 As shown, step S2 includes: S21~S25.

[0080] S21: Generate an initial data set through the multi-scale finite element model established in S11, and the data set is used for the pre-training process of the three neural networks. By inputting simulated loads into the excitation monitoring points in the full-bridge scale finite element model in the initial state, the calculated data of the response monitoring points are extracted as the training data set FR of the neural network FR. If the bridge structure is judged to be intact, the non-monitoring point response signal of the neural network RR data set is obtained by inputting the simulated load into the full-bridge scale finite element model in the initial state; if the bridge structure is judged to be damaged, its non-monitoring point response signal is obtained by inputting the simulated load into the modified full-bridge scale finite element model. According to different bridge stage I assessment states, the corresponding node response data is extracted to obtain the monitoring point response-unmonitored point response training set; based on the model of the multi-scale finite element model, the excitation sensor data and response sensor data in each finite element model node are integrated to obtain the FR-FE training set.

[0081] S22: The generated data set FR is used to train a neural network FR, which is used to quickly calculate the monitoring point field sensor excitation-monitoring point response. Based on the full-bridge scale finite element model of the initial state, the model parameters and the excitation monitoring point data are input into the neural network, and the structural response monitoring point calculation response is output.

[0082] S23: The neural network RR is trained by the generated data set RR. The neural network is used to quickly calculate the corresponding monitoring point signal of the field sensor at the monitoring point - the response of the non-monitoring point. Based on the refined finite element model of the previous time step, the model parameters and the monitoring point response are input into the neural network, and the calculated response of all nodes of the structure is output. The purpose is to obtain the non-monitoring point response by expanding the monitoring point response.

[0083] S24: The neural network FR-FE is trained by the generated data set FR-FE. The neural network is used to analyze the correlation between the data based on the excitation-response data pairs of the field sensors at the monitoring points through the neural network, and output the finite element model calculation parameters. Based on the finite element model of the previous time step, the excitation monitoring point signal and the response monitoring point signal are input into the neural network, and the finite element model calculation parameters are output.

[0084] S25: Adjust the finite element model and neural network through real bridge or scaled model tests.

[0085] For the first stage evaluation stage of the structure, step S3 includes: S31~S35.

[0086] S31: The signals collected on-site from the excitation-response signals of the on-site monitoring point sensors are divided into two parts, one part of the monitoring point excitation is used as the input part of the neural network FR, and the other part of the monitoring point response is used for comparison with the output of the neural network FR.

[0087] S32: Input the excitation monitoring point signal in S31 into the full-bridge finite element model to obtain a neural network FR, and obtain a response monitoring point calculation response signal based on the initial finite element model.

[0088] S33: Compare the calculated response signal of the response monitoring point in S32 with the on-site monitoring response signal in S31, calculate the mean square error between the measured data and the calculated data of all response monitoring points, and determine whether the signal obtained by calculating the full-bridge scale finite element model in the initial state is consistent with the on-site monitoring point signal.

[0089] S34: If the signal comparison error is within the given error range, it means that the offshore bridge structure has not been damaged at the macro level and the macro safety status of the structure is good; if the signal comparison error exceeds the given range, it means that the structure can be judged as damaged at the macro scale based on the on-site sensor signal.

[0090] For step S31, the neural network FR is established by the full-bridge scale finite element model in the initial state of completion and opening to traffic, and is used to quickly calculate the excitation-response of the monitoring points. Because when performing the structural state assessment in stage I, the structure is assumed to be undamaged by default, so the comparison object is the initial undamaged bridge model. If in a certain time step, the stage I assessment determines that the full-bridge scale is damaged, then the subsequent time step does not require the stage I assessment, and directly enters the stage II to locate the damage of the bridge structure and the parameter threshold warning.

[0091] For the second stage evaluation stage of the structure, step S4 includes: S41~S45.

[0092] S41: As in S23, the neural network RR is used, and the response monitoring point signal is input into the neural network RR to obtain the calculated response of all structural nodes of the full bridge scale.

[0093] S42: Calculate the current structural model status assessment parameters according to the warning threshold set in S13 ,according to <2.0%, 2.0%< <5.0%, 5.0%< <20.0%, >20.0% Classify the damage level and locate each non-zero parameter Position, non-zero parameter Corresponding to different damage levels.

[0094] S43: Arrange the structural warning levels in S42 in order of size, and analyze whether the response of the most dangerous position of the structure exceeds the given warning range.

[0095] S44: If the state parameter exceeds the warning threshold, it is necessary to locate the damage position and its corresponding damage level; if the state parameter does not exceed the warning range, all nodes of the entire structure are in a safe state.

[0096] For the third stage of structural evaluation, step S5 includes: S51~S55.

[0097] S51: Analyze the field sensor excitation-response signal data set FR-FE through the neural network FR-FE, directly obtain the calculation parameters of the finite element model, and verify the rationality of the correction parameters of the finite element model through the back calculation function in S24.

[0098] S52: The structure is verified according to the manufacturing specifications by directly using the residual load capacity of the finite element calculation model through the modified full-bridge finite element model.

[0099] S53: Extract the refined modeling part with small mesh size and directly calculate the residual bearing capacity of the component. Calculate the maximum bending moment, maximum shear force and maximum tension that the part can withstand according to the stress condition of the component. , , , and compare the relevant regulations in the specifications to complete the evaluation of the residual bearing capacity level of the components.

[0100] S54: Extract the ratio of each calculated parameter to the calculated parameter of the initial finite element component model through the modified component finite element model, that is, adjust the parameter Using the degradation model of the structural materials determined in S12, the extracted Determine the development status of the component in terms of material properties or geometric properties at this time, so as to determine the remaining use status of the material and calculate the overall service life of the component.

[0101] S55: Count the external excitations (including vehicle loads, temperature changes, etc.) that the structure has been subjected to from the previous time period to this time period. Summarize the finite element model information and integrate the bridge safety status assessment report.

[0102] Further, for step S53, the purpose of this case is to determine the residual bearing capacity of the finite element model modified at this time step, and the evaluation objects are the maximum bending moment, maximum shear force and maximum tension force. , , To achieve the same purpose, the principal stress of the unit can be determined by applying a virtual load to the structure, the principal stress can be compared with the material strength, and the warning level of the structure can be evaluated by setting different warning thresholds. Similarly, this step can use different evaluation methods to adjust the calculation process based on the object of safety assessment.

[0103] For step S54, for the vehicle loads in the initial few time steps, since the vehicle load information of the bridge is not stored on the cloud server, the vehicle load used to predict the remaining service life of the bridge can be replaced by using the vehicle load information of other bridges.

[0104] For step S6, in S5, the full-bridge scale finite element model is analyzed to obtain the quantitative evaluation results of each node by analyzing the excitation and response monitoring point data, and the damage information directly obtained by the sensor is still missing. The finite element model is updated by simulating the damage in the refinement part. The perfect neural network update here is based on the finite element model of the damage.

[0105] Furthermore, the damage information contained in the safety status assessment report integrated in S5 is only derived from sensors, and the finite element model does not simulate the damage in the report. Damage simulation is performed in S6.

[0106] Furthermore, the finite element model of S6 not only needs to simulate sensor damage, but also needs to remove the calculation parameter reduction caused by damage based on the finite element model parameters calculated in S5. Therefore, the calculation parameters of the finite element model also need to be adjusted. The specific method is to calculate the influence of each node caused by the damage and subtract the influence of the damage from the model calculation parameters calculated in S5.

[0107] Specifically, the bridge to be evaluated is a physical bridge to be evaluated for building safety. This embodiment does not impose any restrictions on the structure and material properties of the bridge to be evaluated. The bridge to be evaluated includes but is not limited to bridges of various architectural forms and uses such as ancient building bridges and cross-sea bridges. The original design drawing of the bridge is a drawing reflecting various building data of the bridge drawn based on the built bridge or obtained based on the historical building information of the bridge. The original design drawing of the bridge is annotated with building data and building material information. The bridge connection point is the splicing site of multiple bridge body components for the purpose of constructing the bridge body and balancing the load-bearing force.

[0108] Example 3 This embodiment provides a bridge safety early warning system based on the fusion of data model and physical model, which is used to implement any combination of the above embodiments or solutions and methods, including: The first main control module is used to construct a multi-scale finite element model of the bridge structure, calculate the structural response by inputting simulated excitation data into the finite element model, and generate a data set; A second main control module is used to construct a neural network model and use the data set to train the neural network model, wherein the neural network model includes: a neural network FR, a neural network RR, and a neural network FR-FE; The third main control module is used to calculate the on-site monitoring point excitation-monitoring point response using the neural network FR, obtain the calculated response based on the initial state finite element model, and conduct comparative analysis to determine whether the bridge structure is abnormal or damaged; The fourth main control module is used to use the neural network RR to analyze the response information of the monitoring points input in the third main control module, expand the calculated responses of the non-monitoring points, and classify them to determine the damage level and the location of the damage parameter exceeding the predetermined threshold; The fifth main control module is used to analyze the excitation signal-response signal collected on site through the neural network FR-FE, obtain the correction parameters for updating the finite element model, and calculate the overall bearing capacity level of the bridge and the remaining service life of the structure based on the updated multi-scale finite element model.

[0109] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A bridge safety early warning method based on the fusion of data model and physical model, characterized in that: The following steps are involved: Step S1, constructing a multi-scale finite element model of the bridge structure, calculating the structural response by inputting simulated excitation data into the finite element model, and generating a data set including an excitation signal, a response signal and structural model information; Step S2, constructing a neural network model, and using corresponding data sets to train the neural network model according to different neural network calculation purposes, the neural network model includes: neural network FR, neural network RR and neural network FR-FE; Step S3, first safety assessment stage: using the neural network FR to calculate the on-site monitoring point excitation-monitoring point response, and obtain the calculated response based on the initial state finite element model to determine whether the bridge structure is abnormal or damaged; Step S4, second safety assessment stage: using neural network RR to analyze the monitoring point response information input in step S3, expanding the non-monitoring point calculation response, and performing calculation response monitoring threshold level classification to determine the location and damage level of the component defined as damaged; Step S5, the third safety assessment stage: the excitation signal-response signal collected on site is analyzed by the neural network FR-FE to obtain the correction parameters for the updated finite element model, and the overall bearing capacity level of the bridge and the remaining service life of the structure are calculated based on the updated multi-scale finite element model.

2. A bridge safety early warning method based on the fusion of data model and physical model according to claim 1, characterized in that: Step S1 includes the following steps: Step S11, dividing the construction section where a refined model needs to be established according to the most unfavorable position of the bridge and the position of the pre-buried sensors on site; Step S12, dividing the finite element structure into multiple components according to the sensor data type and layout points, establishing component scale and full bridge scale finite element models according to the characteristic information provided by the engineering drawings, using different grid scales in the same finite element model to simulate monitoring components and non-monitoring components, and establishing a multi-scale finite element model; Step S13, using a multi-scale finite element model to simulate and analyze the main loads and performance changes of the bridge, and establishing a data set by exporting the calculation information and excitation information of different positions of the initial structural finite element model.

3. A bridge safety early warning method based on the fusion of data model and physical model according to claim 2, characterized in that: In step S13, the structural response is calculated by inputting simulated excitation data into the finite element model excitation monitoring point, that is, a suitable degradation model is selected for different structural models, a bridge geometric parameter degradation model is established, and the degradation range of the finite element model calculation parameters compared with the initial model is determined.

4. The bridge safety early warning method based on the fusion of data model and physical model according to claim 1 is characterized in that: In step S2, by inputting simulated loads to the excitation monitoring points in the full-bridge scale finite element model in the initial state, the calculated data of the response monitoring points are extracted as a training data set for the neural network FR; If the bridge structure is judged to be intact, the non-monitoring point response signals of the neural network RR data set are obtained by inputting the simulated load into the full-bridge scale finite element model in the initial state; if the bridge structure is judged to be damaged, its non-monitoring point response signals are obtained by inputting the simulated load into the revised full-bridge scale finite element model, and according to the assessment status of the first safety assessment stage, the corresponding node response data are extracted to obtain the monitoring point response-unmonitoring point response training set; Based on the model of the multi-scale finite element model, the excitation sensor data and response sensor data in each finite element model node are integrated to obtain the FR-FE training set.

5. The bridge safety early warning method based on the fusion of data model and physical model according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31, dividing the signals collected on-site from the excitation-response signals of the on-site monitoring point sensors into two parts, one part of the monitoring point excitation is used as a training process of the neural network FR, and the other part of the monitoring point response is used as a verification set with the neural network FR; Step S32, inputting the excitation monitoring point signal in step S31 into the full-bridge finite element model to obtain a neural network FR, and obtaining a response monitoring point calculation response signal based on the initial finite element model; Step S33, comparing the calculated response signal of the response monitoring point in step S32 with the on-site monitoring response signal in step S31, calculating the mean square error between the measured data and the calculated data of all the response monitoring points, and determining whether the signal obtained by calculating the initial state full-bridge scale finite element model of the verification signal is consistent with the on-site monitoring point signal; Step S34: If the signal comparison error is within the given error range, it means that the offshore bridge structure has not been damaged at a macroscopic level, and the macroscopic safety status of the structure is good; if the signal comparison error exceeds the given range, it means that the structure can be judged as damaged at a macroscopic level through the calculation of the field sensor signal by the neural network FR; Step S35 uses the neural network FR to calculate the on-site monitoring point excitation-monitoring point response, and obtains the calculated response based on the initial state finite element model to determine whether the bridge structure is abnormal or damaged.

6. The bridge safety early warning method based on the fusion of data model and physical model according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41, calculate the current structural model state assessment parameter according to the calculation response of all structural nodes of the full bridge scale obtained by the neural network RR and the set warning threshold , and evaluate the state parameters The damage level is divided according to the value of The threshold value of Step S42: Arrange the structural damage levels in step S41 according to their magnitude, and analyze the state assessment parameters of the most dangerous position of the structure. Whether it exceeds the given warning threshold range; Step S43: If the status evaluation parameter If the warning threshold is exceeded, the damage position and its corresponding damage level need to be located; otherwise, if all node status evaluation parameters If the warning threshold is not exceeded, all nodes in the entire structure are in a safe state.

7. The bridge safety early warning method based on the fusion of data model and physical model according to claim 6 is characterized in that: In step S41, the state evaluation parameter Divided into: <2.0%, 2.0%< <5.0%, 5.0%< <20.0%, >20.0% four levels.

8. The bridge safety early warning method based on the fusion of data model and physical model according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51, input the monitoring point excitation-monitoring point response signal of the field sensor into the neural network FR-FE, and extract the weight matrix of the hidden layer of the neural network; Step S52, based on the inversion function in step S2, calculating the calculation parameters of the multi-scale finite element model, correcting the finite element model in real time, using the corrected multi-scale finite element model to calculate the residual bearing capacity of the structure, and evaluating the overall safety information of the structure; Step S53, calculating the material degradation index at each node based on the modified vector finite element model parameters and the initial vector finite element model parameters , structural cross-sectional area degradation index and the degradation index of the moment of inertia of the structural section According to the on-site environmental factors data obtained by on-site sensor monitoring, the degradation model of material properties in the degradation index is derived, and the material degradation index is used to determine the degradation model of the material properties in the degradation index. Predict the remaining useful life of materials, fit , With the time-varying curve, the correlation between the reduction of the cross-sectional properties and time is derived to obtain the degradation function of the cross-sectional properties; Step S54, evaluating the refined model part of the finite element model according to the material degradation index , structural cross-sectional area degradation index and the degradation index of the moment of inertia of the structural section The finite element component scale model is modified, and the component safety information is evaluated and the residual bearing capacity of the component is calculated based on the finite element model calculation results.

9. The bridge safety early warning method based on the fusion of data model and physical model according to claim 1 is characterized in that: The following steps are also included: The finite element model is directly corrected using the structural damage information directly monitored by the on-site sensors in S3. That is, the final correction of the finite element model needs to be based on the consistency between the structural damage and the on-site monitoring damage. The finite element model is updated according to the degradation coefficients of the section and material analyzed by the monitoring signal in S5 and the adjustment parameters of the corrected structural damage. The updated finite element structural model is used to generate a new data set for training the neural network model and correcting the finite element model.

10. A bridge safety early warning system based on the fusion of data model and physical model, characterized in that: include: The first main control module is used to construct a multi-scale finite element model of the bridge structure, calculate the structural response by inputting simulated excitation data into the finite element model, and generate a data set; A second main control module is used to construct a neural network model and use the data set to train the neural network model, wherein the neural network model includes: a neural network FR, a neural network RR, and a neural network FR-FE; The third main control module is used to calculate the on-site monitoring point excitation-monitoring point response using the neural network FR, obtain the calculated response based on the initial state finite element model, and conduct comparative analysis to determine whether the bridge structure is abnormal or damaged; The fourth main control module is used to use the neural network RR to analyze the response information of the monitoring points input in the third main control module, expand the calculated responses of the non-monitoring points, and classify them to determine the damage level and the location of the damage parameter exceeding the predetermined threshold; The fifth main control module is used to analyze the excitation signal-response signal collected on site through the neural network FR-FE, obtain the correction parameters for updating the finite element model, and calculate the overall bearing capacity level of the bridge and the remaining service life of the structure based on the updated multi-scale finite element model.

Images