A bridge vibration identification method and system based on artificial intelligence technology

Abstract

The present invention discloses a bridge vibration identification method and system based on artificial intelligence technology, which relates to the field of data identification technology. The bridge vibration identification method divides the bridge into multiple bridge sections, collects bridge section vibration data within a preset time, forms a vibration data set, analyzes the stress bearing capacity of the bridge section based on the data, identifies the stress bearing limit of each bridge section, establishes the health and safety threshold of the bridge section, provides indicators for the identification of dangerous bridge sections, inputs real-time monitoring data into a vibration identification model, predicts the degree of vibration impact, dynamically monitors the bridge status, discovers risks in advance and takes measures. Analyze vibration influencing factors, clarify the source of impact, and help optimize response strategies. Predict vibration deformation trends, evaluate bridge structural changes, avoid major accidents, reduce overload stress based on a balance strategy, extend the life of the bridge, regularly monitor the health status, formulate optimized operation time and mechanism, and ensure safe operation of the bridge.

Description

Technical Field

[0001] The present invention relates to the technical field of data recognition, and in particular to a bridge vibration recognition method and system based on artificial intelligence technology. Background Art

[0002] Bridge structures will suffer varying degrees of damage during construction and operation, which shortens their service life. Bridge structure damage identification has become an important factor in ensuring healthy operation during its life cycle. As an indispensable part of civil engineering structures, the safety of bridge structures is of paramount importance. Therefore, it is necessary to identify the degree of vibration damage to existing bridges during operation and obtain the health status of multiple bridge sections. In recent years, most methods use vehicle response to identify the degree of bridge vibration damage, but the vehicle-bridge coupling values ​​are not precise enough and the different stress tolerances and external influencing factors of the bridge sections are not considered, resulting in different degrees of bridge deformation, and thus different risks of cracks or damage.

[0003] In summary, how to identify the vibration health status and deformation degree of the bridge according to the stress tolerance of different bridge sections and combining multi-source data is an urgent problem to be solved and optimized in the bridge vibration identification system based on artificial intelligence technology. Summary of the invention

[0004] The present invention provides a bridge vibration identification method and system based on artificial intelligence technology, which solves the technical problem of how to identify the vibration health status and deformation degree of the bridge according to the stress tolerance of different bridge sections in combination with multi-source data.

[0005] In order to solve the above technical problems, the present invention provides a bridge vibration identification method and system based on artificial intelligence technology. The specific technical solution is as follows:

[0006] In a first aspect, a bridge vibration identification method based on artificial intelligence technology comprises the following steps:

[0007] Dividing the bridge to be tested into a plurality of bridge sections, collecting vibration data of the plurality of bridge sections within a preset time period, so as to obtain a bridge section vibration data set;

[0008] Based on the bridge section vibration data set, the vibration stress tolerance of multiple bridge sections is obtained; based on the vibration stress tolerance, the vibration stress tolerance limit values ​​of multiple bridge sections are obtained; based on the bridge section vibration stress tolerance limit values, the vibration influence degrees of multiple bridge sections are obtained;

[0009] The vibration data set of the bridge section collected within a preset time and a plurality of vibration stress tolerance limits of the bridge section are trained to construct a vibration recognition model, and the vibration recognition model outputs an identification result representing the vibration influence degree of the bridge section; at least one data item in the real-time monitored bridge section vibration data set is input into the vibration recognition model to output a prediction result of the vibration influence degree of the bridge section;

[0010] Based on the vibration influence degree of the bridge section, a vibration influence factor is obtained; based on the vibration influence factor, the influence property of the vibration influence factor is determined; according to the influence property, a vibration deformation trend of the bridge section is obtained; based on the deformation trend, a bridge section stress tolerance balance strategy is obtained to balance the vibration stress tolerance of the bridge section;

[0011] Based on the balanced bridge section vibration stress tolerance, the bridge section stress tolerance health status is obtained regularly; and according to the bridge section stress tolerance health status, the bridge section operation time and operation mechanism are planned.

[0012] As a further optimization scheme of the present invention, based on the bridge segment vibration data set, the vibration stress tolerance of multiple bridge segments is obtained, including:

[0013] Preprocessing the bridge section vibration data set to obtain a preprocessed data set; extracting vibration features from the preprocessed data set to obtain a feature data set; the feature data set includes time domain feature data and frequency domain feature data;

[0014] According to the characteristic data set, the vibration stress change of the bridge section is obtained; based on the vibration stress change of the bridge section, , to obtain the dynamic strain, where represents the dynamic strain, represents the stress variation coefficient, L represents the length of the bridge section, and E represents the dynamic modulus of the vibration signal;

[0015] Based on the dynamic strain, , so that the dynamic strain is converted into a vibration stress value; the vibration stress is in a dynamic change state, and the vibration stress value is converted into a vibration stress value through

[0016] , so that the vibration stress value is converted into an equivalent stress scalar, which represents a comparable stress value, where, represents the vibration stress value, e represents the conversion parameter, represents the equivalent stress scalar, T represents the frequency period of the vibration signal, and dt represents the period increment;

[0017] Based on the equivalent stress scalar, , to obtain the fatigue bearing capacity; the fatigue bearing capacity is the vibration stress bearing capacity of the bridge section, where, It represents the vibration stress tolerance of the bridge section, m represents the material coefficient of the bridge section, and K represents the fatigue constant.

[0018] As a further optimization scheme of the present invention, according to the vibration stress tolerance, multiple bridge section vibration stress tolerance limits are obtained, including:

[0019] Based on the vibration stress tolerance, To obtain the fatigue limit strength of the bridge section; the fatigue limit strength of the bridge section is the vibration stress tolerance limit of the bridge section, where: Indicates the fatigue limit strength of the bridge segment;

[0020] Based on the vibration stress tolerance limit of the bridge section, a phased assessment of the health status of the bridge section is performed; when the vibration stress tolerance limit of the bridge section >1, indicating that the bridge section is in a healthy state; when the vibration stress tolerance of the bridge section ≤1, indicating that the bridge segment has an imbalance risk;

[0021] Based on the health status of the bridge segment, The healthy bridge section is checked to obtain the bridge section vibration stress tolerance correction limit; the bridge section vibration stress tolerance correction limit is used to reduce the error risk of the healthy state of the bridge section; where, Indicates the modified limit of the vibration stress of the bridge section. represents the preset fatigue limit value, k represents the correction parameter, and m represents the fatigue parameter.

[0022] As a further optimization scheme of the present invention, the vibration influence degrees of multiple bridge sections are obtained according to the vibration stress tolerance limit of the bridge section, including:

[0023] According to the health status of the bridge segment, ; To obtain the vibration influence of the bridge section; The vibration influence of the bridge section is used to measure the influence of vibration on the health status of the bridge section; Where, It indicates the influence degree of bridge section vibration, and z indicates the normalized index of influence degree;

[0024] Based on the influence degree of the bridge section vibration, the influence results of the health status of the multiple bridge sections on the vibration are judged; the influence results include Indicates that the amplitude and frequency of the vibration signal are small and have no significant impact on the bridge structure; when Indicates that the vibration signal amplitude and frequency are close to the limit and need to be closely monitored; It means that the amplitude and frequency of the dynamic signal exceed the limit value, there is a risk of bridge section damage or fatigue, and repair measures need to be taken.

[0025] As a further optimization solution of the present invention, the vibration recognition model includes:

[0026] Generate structural data from the bridge section vibration data set and a plurality of bridge section vibration stress tolerance limits within a preset time, and encode the structural data into sequence data to train and obtain the vibration recognition model;

[0027] The sequence data is input into the vibration recognition model; the vibration recognition model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, and the intermediate representation data of multiple hidden layers are transmitted to the output layer, and the output layer output represents the recognition result of the vibration influence degree of the bridge section;

[0028] At least one data item of the newly acquired bridge section vibration data set is input into the vibration recognition model, and the output layer outputs the prediction result of the bridge section vibration influence degree representing the newly acquired bridge section vibration data set.

[0029] As a further optimization scheme of the present invention, based on the vibration influence degree of the bridge section, a vibration influence factor is obtained, including:

[0030] The result of the influence of the vibration on the health status of the bridge section is obtained according to the influence degree of the vibration of the bridge section; multiple bridge sections are marked to obtain a structure set ; In the formula, n represents the impact factor category;

[0031] According to the structure set, the bridge load weight is determined ,pass , to obtain the bridge section influence factor load coefficient; the bridge section influence factor load coefficient represents the degree of influence on the bridge section vibration aggravation; where, represents the load coefficient of the bridge section impact factor, Indicates the influence degree of bridge section vibration of the nth category influencing factor;

[0032] Based on the load factor of the bridge section, To obtain the load coefficient of the entire bridge impact factor.

[0033] As a further optimization scheme of the present invention, based on the vibration influence factor, the influence property of the vibration influence factor is judged, including:

[0034] According to the load coefficient of the whole bridge influencing factor, the influencing nature of the vibration influencing factor is judged; the influencing nature of the vibration influencing factor includes direct influence and indirect influence;

[0035] Set the impact factor load threshold, when the full bridge impact factor load coefficient is greater than the impact factor load threshold, then The attribute is directly affected; otherwise, The property is an indirect influence.

[0036] As a further optimization scheme of the present invention, the vibration deformation trend of the bridge section is obtained according to the influencing properties; and the stress tolerance balance strategy of the bridge section is obtained based on the deformation trend to balance the vibration stress tolerance of the bridge section, including:

[0037] Based on the influencing properties of the vibration influencing factors, vibration response data is obtained; according to the vibration response data, the vibration signal mode is decomposed to obtain a plurality of modal parameters;

[0038] According to multiple modal parameters, , to obtain the bridge segment deformation excitation vector; where, represents the bridge segment deformation excitation vector, K represents the stiffness matrix, C represents the damping matrix, M represents the mass matrix, and u(t) represents the displacement vector; according to the bridge segment deformation excitation vector, the vibration displacement change trend of the bridge segment is reflected, and the vibration displacement change trend is the vibration deformation trend of the bridge; when multiple modal parameters are within the preset threshold range, it indicates that the bridge segment deformation is stable or there is no deformation phenomenon;

[0039] According to the bridge segment deformation excitation vector, the bridge segment deformation excitation vector is encoded into sequence data to train and construct a bridge segment dynamic deformation trend model; at least one modal parameter in the bridge segment deformation excitation vector at the i-th time node is input into the bridge segment dynamic deformation trend model to output the bridge segment vibration deformation trend at the next time node;

[0040] Based on the output results of the dynamic deformation trend model of the bridge section, the internal stress distribution data of the bridge section is obtained through the established target optimization function and preset constraints; based on the internal stress distribution data of the bridge section, the stress bearing balance strategy of the bridge section is obtained.

[0041] As a further optimization solution of the present invention, the bridge section stress bearing balance strategy includes:

[0042] According to the multiple modal parameters, deformation stress analysis is performed on the bridge segment to obtain bridge segment stress distribution data; based on the bridge segment stress distribution data, deformation of the bridge segment is graded and labeled to obtain graded labeled data; the graded labeled data includes a first-order deformation bridge segment, a second-order deformation bridge segment and a third-order deformation bridge segment;

[0043] Based on the graded and labeled data, the load of the bridge is distributed so that the deformation stress of each bridge section tends to a stable state, so as to alleviate the stress fluctuation caused by vibration, and the coordinated bearing capacity between the bridge sections is improved through reinforcement design; the stress bearing health status of multiple graded and labeled bridge sections is obtained regularly; according to the stress bearing health status of the bridge sections, the operation time and operation mechanism of the bridge sections are planned.

[0044] In a second aspect, the system is provided with an electronic device including a memory, a processor, and a bridge vibration identification method program based on artificial intelligence technology stored in the memory and executable on the processor, wherein the bridge vibration identification method program based on artificial intelligence technology implements the steps of a bridge vibration identification method based on artificial intelligence technology when executed by the processor, and the system includes:

[0045] Data acquisition module: used to divide the bridge to be tested into multiple bridge sections, collect vibration data of the multiple bridge sections within a preset time, so as to obtain a bridge section vibration data set;

[0046] Data identification module: used to obtain the vibration stress tolerance of multiple bridge sections based on the bridge section vibration data set; obtain the vibration stress tolerance limit values ​​of multiple bridge sections according to the vibration stress tolerance values; and obtain the vibration influence degrees of multiple bridge sections according to the vibration stress tolerance limit values ​​of the bridge sections;

[0047] Data processing module: it is used to obtain the vibration influence factor based on the vibration influence degree of the bridge section; to judge the influence property of the vibration influence factor based on the vibration influence factor; to obtain the vibration deformation trend of the bridge section according to the influence property; to obtain the bridge section stress tolerance balance strategy based on the deformation trend to balance the vibration stress tolerance of the bridge section;

[0048] Operation planning module: It is used to regularly obtain the stress tolerance health status of the bridge section based on the balanced bridge section vibration stress tolerance; and plan the operation time and operation mechanism of the bridge section according to the stress tolerance health status of the bridge section.

[0049] The present invention has at least the following beneficial effects: by dividing the bridge into multiple bridge sections, the present invention can refine the monitoring scope, facilitate the positioning of problem areas, and improve monitoring accuracy; collect vibration data within a preset time to ensure that the data has a time span and dynamics, reflecting the actual vibration state of the bridge under different time and load conditions; form a bridge section vibration data set, laying a comprehensive data foundation for subsequent analysis.

[0050] Based on the bridge section vibration data set, the vibration stress tolerance of multiple bridge sections is obtained; the stress tolerance of the bridge section is analyzed through the vibration data to clarify the actual load-bearing state of each bridge section; this helps to discover the weaknesses of the bridge under vibration conditions and improve the pertinence of the bridge safety assessment;

[0051] According to the vibration stress tolerance, the vibration stress tolerance limits of multiple bridge sections are obtained: the ultimate bearing capacity (vibration stress tolerance limit) of the bridge section is extracted, and the health and safety threshold range of the bridge section is established; important indicators are provided for the subsequent identification of dangerous bridge sections, facilitating the timely formulation of response strategies.

[0052] The vibration impact degree of multiple bridge sections is obtained based on the vibration stress tolerance limit of the bridge sections. The vibration impact degree of the bridge sections is evaluated to clarify the potential threat of vibration to the health of the bridge section structure. By quantifying the impact degree, a scientific basis is provided for maintenance and operation planning.

[0053] The vibration data set of the bridge section collected within the preset time and the vibration stress tolerance limit training are used to construct a vibration recognition model, and the recognition result of the vibration impact degree of the bridge section is output: the construction of the vibration recognition model can realize the automatic analysis and pattern recognition of the data, and improve work efficiency; the output recognition result can quickly screen the problem area and lay the model foundation for subsequent real-time prediction.

[0054] At least one data item in the real-time monitored bridge section vibration data set is input into the vibration identification model to output the prediction result of the bridge section vibration impact degree: through the real-time data input model, the vibration impact degree of the bridge section is predicted to realize dynamic monitoring of the bridge status; potential risks are discovered in advance and preventive measures are taken to improve the safety management capabilities of the bridge.

[0055] Based on the vibration impact degree of the bridge section, the vibration impact factor is obtained: the impact factor is extracted to identify the main causes of bridge section vibration (such as vehicle traffic load, environmental changes, weather factors and bridge material aging, etc.); it helps to control or optimize the vibration impact from the source and improve the stability of the bridge.

[0056] Based on the vibration impact factor, the nature of the impact of the vibration impact factor is judged: by analyzing the impact nature, it is possible to distinguish whether the vibration impact is short-term sporadic or long-term continuous, providing a decision-making basis for different response measures; accurately judging the nature of vibration can help prioritize critical or high-risk issues.

[0057] Obtain the vibration deformation trend of the bridge section according to the nature of the impact: Predicting the deformation trend caused by the vibration of the bridge section is helpful to evaluate the structural changes of the bridge during long-term use; timely understand whether the bridge has accelerated or abnormal deformation trends to avoid major accidents.

[0058] Based on the deformation trend, the stress balance strategy of the bridge section is obtained to balance the vibration stress tolerance of the bridge section: a balance strategy is proposed to reduce the overload stress of specific bridge sections by guiding traffic, optimizing loads, etc., thereby extending the life of the bridge; and avoiding premature aging or damage of the bridge due to local stress concentration.

[0059] Based on the balanced vibration stress tolerance of the bridge section, the stress tolerance health status of the bridge section is obtained regularly: the health status is monitored regularly, a bridge health file is formed, and the use status of the bridge is dynamically understood; the scientificity and accuracy of bridge management are improved, and maintenance costs are reduced.

[0060] Plan the operation time and mechanism of the bridge section according to its stress-bearing health status: formulate optimized operation time and mechanism (such as weight limit, traffic limit, speed limit, etc.) based on the health status to ensure that the bridge operates within a safe range; reasonably plan the operation mechanism to improve the utilization efficiency and service life of the bridge. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flow chart of a bridge vibration identification method based on artificial intelligence technology provided by an embodiment of the present invention;

[0062] Figure 2 is a schematic diagram of a bridge vibration identification system based on artificial intelligence technology provided by an embodiment of the present invention;

[0063] Figure 3 is a first-order deformation bridge segment model diagram provided by an embodiment of the present invention;

[0064] Figure 4 is a diagram of a second-order deformation bridge segment model provided by an embodiment of the present invention;

[0065] Figure 5 It is a diagram of a three-order deformation bridge segment model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.

[0067] This embodiment provides a bridge vibration identification method and system based on artificial intelligence technology, and the specific implementation methods are as follows:

[0068] like Figure 1 As shown, a bridge vibration identification method based on artificial intelligence technology includes the following steps:

[0069] Step 11, dividing the bridge to be tested into a plurality of bridge sections, collecting vibration data of the plurality of bridge sections within a preset time, so as to obtain a bridge section vibration data set;

[0070] Step 12, based on the bridge section vibration data set, obtain the vibration stress tolerance of multiple bridge sections; according to the vibration stress tolerance, obtain the vibration stress tolerance limit values ​​of multiple bridge sections; according to the bridge section vibration stress tolerance limit values, obtain the vibration influence degrees of multiple bridge sections;

[0071] The vibration data set of the bridge section collected within a preset time and a plurality of vibration stress tolerance limits of the bridge section are trained to construct a vibration recognition model, and the vibration recognition model outputs an identification result representing the vibration influence degree of the bridge section; at least one data item in the real-time monitored bridge section vibration data set is input into the vibration recognition model to output a prediction result of the vibration influence degree of the bridge section;

[0072] Step 13, based on the vibration influence degree of the bridge section, a vibration influence factor is obtained; based on the vibration influence factor, the influence property of the vibration influence factor is determined; according to the influence property, a vibration deformation trend of the bridge section is obtained; based on the deformation trend, a bridge section stress tolerance balancing strategy is obtained to balance the vibration stress tolerance of the bridge section;

[0073] Step 14: Based on the balanced bridge section vibration stress tolerance, regularly obtain the bridge section stress tolerance health status; and plan the bridge section operation time and operation mechanism according to the bridge section stress tolerance health status.

[0074] In the implementation of the present invention, step 11 divides the bridge to be tested into multiple bridge sections (such as divided by structural function, length or key nodes). In this implementation, it is preferred to divide the bridge into five equal parts, and use the middle symmetry line as the reference line, and the load stress on each symmetrical bridge section is at the same level.

[0075] Within the preset time, the vibration data of each bridge section is collected through the installed vibration sensors, and the Kalman filter algorithm is used to identify the vibration of the bridge structure with the collected bridge section vibration data to form a vibration data set (such as a time series of acceleration, velocity or displacement data); through data analysis, the vibration stress tolerance of each bridge section (that is, the maximum stress range that the bridge can withstand when vibrating) is calculated; the vibration monitoring of the entire bridge is decentralized to facilitate capturing the vibration characteristics of specific locations and improve monitoring accuracy; after obtaining the vibration stress tolerance limit, it can provide a quantitative basis for the health assessment of the bridge; based on the degree of vibration impact, the area with a larger vibration response can be quickly located and focused on monitoring.

[0076] Step 12 inputs the bridge section vibration data set (input) collected within the preset time of step 11 and the vibration stress tolerance limit (label) of each bridge section into the machine learning model to build a vibration recognition model; the output of the vibration recognition model is the recognition result of the vibration impact degree of the bridge section (such as mild, moderate, severe); the real-time monitored vibration data is input into the vibration recognition model to obtain the predicted result of the vibration impact degree of the bridge section; the model is used to automatically analyze the vibration data to reduce manual intervention and improve efficiency; potential risks are quickly identified by real-time data input and output of vibration impact degree prediction; the machine learning model is used to enhance the recognition accuracy and prediction ability of the vibration impact degree to avoid human errors.

[0077] Step 13 is based on the vibration influence degree of the bridge section obtained in step 12, so as to obtain the vibration influence factor (such as vibration frequency, amplitude, stress change rate, etc.) to describe the specific influence of vibration on the bridge section; by analyzing the vibration influence factor, it is judged whether it is beneficial (such as eliminating some stress concentration) or unfavorable (such as inducing crack extension) to the bridge section structure; based on historical and real-time data, the future deformation trend of the bridge section (such as displacement increase, crack extension direction, etc.) is predicted; according to the deformation trend, a stress bearing balance strategy is proposed (such as optimizing vehicle traffic path, adjusting traffic load, and dynamically strengthening local areas); by analyzing the vibration properties, the positive and negative effects of vibration are distinguished to provide a basis for subsequent decision-making; through the balance strategy, the load on the high stress concentration area is reduced to reduce the structural fatigue damage rate; based on data-driven stress bearing strategy optimization, the scientificity and flexibility of bridge maintenance are improved.

[0078] Step 14: Based on the vibration stress tolerance of the bridge section after equalization in step 13, regularly evaluate the health status of the bridge (such as the degree of damage of each part of the bridge and the remaining value of fatigue life); adjust the operation time of the bridge according to the health status (such as night-time load limit or time-sharing traffic); design a reasonable operation strategy (such as optimizing traffic load and restricting the passage of overweight vehicles) based on the health status and use requirements of the bridge; based on the dynamic update of the health status of the bridge section, realize real-time grasp of the overall health status of the bridge; reduce secondary structural damage caused by excessive vibration by optimizing the operation time and mechanism; reasonably arrange the operation plan to reduce the impact of high-load operation on the bridge and improve the long-term stability and safety of the bridge.

[0079] The above steps work together, from data collection to model prediction, and then to health status assessment, covering the entire life cycle of bridge monitoring; based on bridge section level monitoring, the accuracy of problem location and resolution is improved; through model training and real-time prediction, human errors are reduced and the automation level of the system is improved; and stress tolerance is balanced and operation strategies are planned in advance to ensure the long-term safety of the bridge.

[0080] In a preferred embodiment of the present invention, the step 12 further comprises: obtaining the vibration stress tolerance of multiple bridge sections based on the bridge section vibration data set;

[0081] Step 121, preprocessing the bridge section vibration data set to obtain a preprocessed data set; extracting vibration features from the preprocessed data set to obtain a feature data set; the feature data set includes time domain feature data and frequency domain feature data;

[0082] Step 122, obtaining the bridge section vibration stress change according to the characteristic data set; based on the bridge section vibration stress change, , to obtain the dynamic strain, where represents the dynamic strain, represents the stress variation coefficient, L represents the length of the bridge section, and E represents the dynamic modulus of the vibration signal;

[0083] Step 123, based on the dynamic strain, , so that the dynamic strain is converted into a vibration stress value; the vibration stress is in a dynamic change state, and the vibration stress value is converted into a vibration stress value through

[0084] , so that the vibration stress value is converted into an equivalent stress scalar, which represents a comparable stress value, where, represents the vibration stress value, e represents the conversion parameter, represents the equivalent stress scalar, T represents the frequency period of the vibration signal, and dt represents the period increment;

[0085] Step 124, based on the equivalent stress scalar, by , to obtain the fatigue bearing capacity; the fatigue bearing capacity is the vibration stress bearing capacity of the bridge section, where, It represents the vibration stress tolerance of the bridge section, m represents the material coefficient of the bridge section, and K represents the fatigue constant.

[0086] In the implementation of the present invention, step 121 collects the original bridge section vibration data, which usually contains noise, outliers or redundant data. Through preprocessing, these unnecessary interference information can be removed, making the subsequent analysis more accurate. The preprocessing step may include operations such as denoising, filtering, and normalization; after data preprocessing, it is necessary to extract useful feature information from the vibration data. The characteristics of the vibration signal can be divided into time domain characteristics and frequency domain characteristics; by analyzing the time domain waveform of the vibration signal, statistical characteristics such as mean, standard deviation, peak, skewness, kurtosis, etc. can be extracted; through frequency domain analysis methods such as Fourier transform, frequency-related features such as spectrum, resonant frequency, etc. are extracted. These features help to reveal the potential laws and structural health status of bridge vibration; data preprocessing and feature extraction can effectively eliminate interference and improve the accuracy of subsequent analysis; the extracted time domain and frequency domain features can provide key parameters for subsequent stress change analysis and fatigue assessment.

[0087] Step 122 can further analyze the stress changes caused by bridge vibration based on the feature data set extracted in step 121. By calculating the stress changes, the stress changes of the bridge section under vibration can be estimated more accurately. This analysis provides a quantitative basis for fatigue assessment and health monitoring of the bridge.

[0088] Step 123: The dynamic strain obtained in step 122 reflects the stress change trend of the bridge during the vibration process, and is further converted into an actual vibration stress value. Since the vibration stress usually presents a dynamic change state, it is necessary to convert the vibration stress into a stable and easy-to-compare scalar value, namely, an "equivalent stress scalar". Converting the dynamic stress into an equivalent stress scalar makes the vibration stress value comparable, which is convenient for comparative analysis with other materials or structures. Through this conversion, complex dynamic change problems can be simplified into stable scalars that are easy to understand and apply.

[0089] In step 124, the fatigue bearing capacity of the bridge section is calculated based on the equivalent stress scalar obtained in step 123, combined with the material coefficient and fatigue constant of the bridge; thereby providing a quantitative fatigue bearing capacity value for the health assessment of the bridge, which helps to predict the service life of the bridge; by analyzing the fatigue bearing capacity, the potential fatigue damage of the bridge can be effectively identified, providing a basis for subsequent maintenance and reinforcement.

[0090] The core purpose of the above four steps is to evaluate the fatigue resistance of the bridge and predict its remaining service life through the analysis of bridge vibration data; to improve the accuracy and reliability of vibration data analysis; to achieve quantitative evaluation of vibration stress and provide data support for bridge health management; and to help decision makers reasonably arrange bridge maintenance and reinforcement work based on fatigue resistance evaluation to ensure safe operation of the bridge.

[0091] In a preferred embodiment of the present invention, the step 12 further comprises: obtaining the vibration stress tolerance limits of multiple bridge sections according to the vibration stress tolerance;

[0092] Step 125, based on the vibration stress tolerance, To obtain the fatigue limit strength of the bridge section; the fatigue limit strength of the bridge section is the vibration stress tolerance limit of the bridge section, where: Indicates the fatigue limit strength of the bridge segment;

[0093] Step 126, based on the vibration stress tolerance limit of the bridge section, a phased evaluation of the health status of the bridge section is performed; when the vibration stress tolerance limit of the bridge section >1, indicating that the bridge section is in a healthy state; when the vibration stress tolerance of the bridge section ≤1, indicating that the bridge segment has an imbalance risk;

[0094] Step 127, based on the health status of the bridge segment, The healthy bridge section is checked to obtain the bridge section vibration stress tolerance correction limit; the bridge section vibration stress tolerance correction limit is used to reduce the error risk of the healthy state of the bridge section; where, Indicates the modified limit of the vibration stress of the bridge section. represents the preset fatigue limit value, k represents the correction parameter, and m represents the fatigue parameter.

[0095] In the implementation of the present invention, the vibration stress tolerance in step 125 refers to the maximum stress level that the bridge section can withstand during the vibration process. This stress level is usually determined by various factors such as the material properties, structural design, and usage of the bridge; the fatigue limit strength refers to the maximum stress value that the bridge material or structure can withstand under long-term cyclic loads without fatigue damage.

[0096] By analyzing the vibration response of the bridge section and combining the characteristics of the structure and materials, its corresponding fatigue limit strength can be calculated; by accurately obtaining the fatigue limit strength, it can provide an accurate benchmark for subsequent bridge section health assessment and avoid bridge failure due to fatigue damage; thereby helping to determine the relationship between vibration stress tolerance and fatigue limit, so that the durability of the bridge can be effectively evaluated.

[0097] Step 126 evaluates the health status of the bridge section according to the vibration stress tolerance limit of the bridge section obtained in step 125 (i.e., the fatigue limit strength of the bridge section); when the vibration stress tolerance of the bridge section is greater than 1, it means that the vibration stress that the bridge section can withstand is greater than the current actual stress, that is, the bridge section is in a healthy state and has no significant risk of damage; when the vibration stress tolerance of the bridge section is less than or equal to 1, it means that the bridge section is in a stress over-limit or critical state and may have an imbalance risk, and further evaluation and reinforcement should be carried out at this time; by evaluating the health status of the bridge section, potential fatigue problems can be discovered in time to avoid accidents caused by excessive fatigue damage to the bridge; using vibration stress tolerance as a standard for determining the health status can realize data-based dynamic monitoring and evaluation, thereby improving the efficiency and accuracy of bridge health management.

[0098] Step 127 After evaluating the health status, the healthy bridge section is checked to obtain a more accurate correction limit of the vibration stress of the bridge section; the correction limit is adjusted to the original limit by considering possible error factors (such as environment, use conditions, material degradation, etc.) in combination with the preset fatigue limit value, correction parameter k and fatigue parameter m; wherein the fatigue limit value is the fatigue limit of the bridge section material set according to historical experience or experimental data, and the correction parameter k and fatigue parameter m are used to reflect the actual working conditions and material properties of the specific bridge section; the obtained correction limit can improve the assessment accuracy of the fatigue condition of the bridge section, especially in the face of complex or changing working conditions, which helps to reduce the health status assessment errors caused by errors, thereby providing a more conservative and reliable assessment basis for the long-term health management of the bridge and enhancing the safety of the bridge.

[0099] The above steps work together to ensure the accuracy of the assessment by obtaining the fatigue limit strength of the bridge segment as the basis. Then, through the health status assessment based on vibration stress tolerance, potential risks are discovered in a timely manner to prevent accidents. Finally, through the correction limit verification, errors are reduced, the assessment accuracy is improved, and the long-term safety of the bridge is further enhanced.

[0100] In a preferred embodiment of the present invention, the step 12 further includes obtaining the vibration influence degrees of multiple bridge sections according to the vibration stress tolerance limit of the bridge section:

[0101] Step 128, according to the health status of the bridge segment, ; To obtain the vibration influence of the bridge section; The vibration influence of the bridge section is used to measure the influence of vibration on the health status of the bridge section; Where, It indicates the influence degree of bridge section vibration, and z indicates the normalized index of influence degree;

[0102] Step 129, based on the vibration influence of the bridge section, determine the influence of the vibration on the health status of the multiple bridge sections; the influence result includes Indicates that the amplitude and frequency of the vibration signal are small and have no significant impact on the bridge structure; Indicates that the vibration signal amplitude and frequency are close to the limit and need to be closely monitored; It means that the amplitude and frequency of the dynamic signal exceed the limit value, there is a risk of bridge section damage or fatigue, and repair measures need to be taken.

[0103] In the implementation of the present invention, step 128 reflects the comprehensive impact of the vibration signal (such as amplitude and frequency) on the health status of the bridge section through the vibration impact degree, which is used to quantify the potential harm of vibration to the bridge section structure; the impact degree normalization index represents the index after the normalization of the amplitude, frequency, vibration mode and other parameters of the vibration signal, which is used to eliminate the parameter differences between different bridge sections; the result is obtained through the health status of the bridge section, which is determined in combination with the vibration parameters and the structural characteristics of the bridge section (such as stress distribution and material fatigue coefficient); the frequency and amplitude of the bridge section vibration are collected by the sensor; the collected vibration signal parameters are normalized to eliminate the differences between different bridge sections (determination of the normalization index); the vibration impact degree is calculated by combining the health status and the normalized parameters; the complex vibration characteristics (amplitude, frequency, etc.) are converted into a clear quantitative index, which is convenient for directly measuring the impact of vibration on the health status of the bridge section; basic data is provided for subsequent bridge section health assessment and maintenance decisions, especially when the vibration may cause fatigue damage, it can be intuitively judged whether intervention measures need to be taken; as the vibration signal changes, the impact of vibration on the health of the bridge section can be tracked in real time, providing dynamic support for bridge maintenance.

[0104] Step 129 classifies the impact results of the health status of the bridge section according to the value of the vibration impact degree:

[0105] It indicates that the amplitude and frequency of the vibration signal are low, and there is no significant impact on the bridge section structure, and the health status of the bridge section is stable; Indicates that the amplitude and frequency of the vibration signal are close to the design or fatigue limit. At this time, the bridge section needs to be closely monitored to prevent fatigue accumulation from causing damage; It means that the amplitude and frequency of the vibration signal exceed the design limit, indicating that the bridge section has been significantly damaged or there is a fatigue risk, and repair or reinforcement measures are needed; the vibration impact degree of each bridge section is calculated, and the health status of the entire bridge is judged after summary analysis; when the vibration impact degree value of some bridge sections is low, the bridge as a whole is in a safe state; if some bridge sections exceed the limit, targeted repairs are required.

[0106] By classifying the different degrees of vibration impact, we can clarify the response measures under different health conditions and ensure the efficient use of resources; specifically identify and manage bridge sections that are most affected by vibration to avoid large-scale repairs or misjudgment of health status; for bridge sections close to the limit, we can intervene early, conduct close monitoring or small-scale reinforcement to prevent further damage to the bridge section and extend the overall life of the bridge; when the degree of vibration impact exceeds the limit, we can quickly take repair measures to prevent major safety accidents caused by structural damage to the bridge.

[0107] In a preferred embodiment of the present invention, the vibration recognition model in step 12 further includes:

[0108] Generate structural data from the bridge section vibration data set and a plurality of bridge section vibration stress tolerance limits within a preset time, and encode the structural data into sequence data to train and obtain the vibration recognition model;

[0109] The sequence data is input into the vibration recognition model; the vibration recognition model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, and the intermediate representation data of multiple hidden layers are transmitted to the output layer, and the output layer output represents the recognition result of the vibration influence degree of the bridge section;

[0110] At least one data item of the newly acquired bridge section vibration data set is input into the vibration recognition model, and the output layer outputs the prediction result of the bridge section vibration influence degree representing the newly acquired bridge section vibration data set.

[0111] In the implementation of the present invention, vibration data is collected from the sensors of the bridge section, including signal characteristics such as vibration amplitude and frequency within a preset time period; the vibration limit defined by the material properties, design specifications, etc. of the bridge section indicates the safety limit of the bridge section when subjected to vibration; the vibration data set is combined with the vibration stress tolerance limit, and structural data of the bridge section vibration is generated through feature extraction or formula processing (such as normalization, difference calculation), reflecting the current vibration state of the bridge section.

[0112] For example: after normalizing the amplitude and frequency of the vibration signal, compare them with the stress limit to generate new characteristic indicators (such as residual safety margin, frequency deviation); encode the generated structural data into sequence data in chronological order or bridge sequence as model training input. For example, the time series of vibration amplitude, frequency characteristics, etc. are formed into a unified input vector. The vibration recognition model is a multi-layer neural network, which specifically includes:

[0113] Input layer: receives the vibration sequence data of the bridge segment.

[0114] Hidden layer: The model contains multiple hidden layers (first, second, and third hidden layers), and each layer extracts deeper features through nonlinear activation functions.

[0115] The first hidden layer: extracts basic features, such as the amplitude change of the vibration signal.

[0116] The second hidden layer further captures complex features, such as the association between frequency features and stress limits.

[0117] The third hidden layer: forms comprehensive features in high-dimensional space to represent the healthy state of bridge segment vibration.

[0118] Output layer: Combined with the intermediate representation data of the hidden layer, the recognition result is output to indicate the vibration impact degree of the bridge section.

[0119] The generated sequence data is used as training samples, combined with the corresponding vibration impact degree labels (such as the classification of vibration impact on bridge section health: no impact, need to be monitored, need to be repaired); by optimizing the loss function, the model parameters (such as weights and biases) are adjusted to make the output results as close to the true labels as possible; at least one data item of the newly acquired vibration data set (such as real-time vibration amplitude and frequency series) is input into the model; the vibration recognition model generates intermediate representation data of the hidden layer based on the new data, and outputs the prediction results through the output layer; the output result is the predicted value of the vibration impact degree, reflecting the vibration health status of the current bridge section (such as safe, need to be monitored or need to be repaired); by combining the vibration data with the stress tolerance limit, the structural data is generated and encoded into sequence data, which provides a clear basis for the quantitative modeling of vibration impact; the encoding method is compatible with the complexity of different bridge sections and unified into the model input format to facilitate model training and application.

[0120] The vibration recognition model extracts the deep relationship of vibration characteristics through multiple hidden layers, which helps to identify complex vibration patterns (such as frequency coupling effects); the intermediate representation data of the hidden layer gradually transforms the basic characteristics of the vibration signal into high-level comprehensive features, improving the recognition accuracy of the model; after training, the vibration recognition model can quickly process newly acquired vibration data sets and output vibration impact prediction results in real time; it supports dynamic monitoring of the health status of bridge sections and timely identification of potential risks.

[0121] The output results of the model can directly quantify the degree of vibration impact on the bridge section, making it easier for engineering personnel to quickly determine the health status of the bridge section. Different health statuses (safe, requiring monitoring, requiring repair) can be classified and managed to optimize the allocation of bridge maintenance resources. By continuously inputting real-time vibration data, the recognition model can dynamically adjust the judgment of the health status of the bridge section, supporting health monitoring and maintenance decisions of the bridge throughout its life cycle.

[0122] In a preferred embodiment of the present invention, the step 13 further comprises: obtaining a vibration influence factor based on the vibration influence degree of the bridge section;

[0123] Step 131, based on the vibration influence degree of the bridge segment, the bridge segment health status is affected by the vibration; multiple bridge segments are marked to obtain a structure set ; In the formula, n represents the impact factor category;

[0124] Step 132, determining the bridge section load weight according to the structure set ,pass , to obtain the bridge section influence factor load coefficient; the bridge section influence factor load coefficient represents the degree of influence on the bridge section vibration aggravation; where, represents the load coefficient of the bridge section impact factor, Indicates the influence degree of bridge section vibration of the nth category influencing factor;

[0125] Step 133: Based on the bridge section impact factor load coefficient, To obtain the load coefficient of the entire bridge impact factor.

[0126] In the implementation of the present invention, step 131 classifies the effect of vibration on the health status of each bridge section according to the vibration effect degree of the bridge section output by the vibration identification model (such as no effect, slight effect, severe effect); the classification results are recorded in a marking manner to form a "structure set" containing multiple bridge sections, which describes the vibration effect status of different bridge sections of the entire bridge;

[0127] Health status = no impact, the corresponding mark is 0;

[0128] Health status = monitoring required, corresponding mark is 1;

[0129] Health status = needs repair, the corresponding mark is 2.

[0130] The categories of influencing factors can be different characteristic factors such as bridge section material, bridge section location (main bridge section / secondary bridge section), vibration frequency, temperature change, etc. The vibration impact degree of each category of influencing factors is evaluated separately through the model; by marking the vibration impact of multiple bridge sections, a structure set is formed to provide basic data support for subsequent calculations; the health status of bridge sections is classified into categories, which helps to optimize the resource allocation of bridge maintenance (such as giving priority to repairing severely affected bridge sections); by introducing influencing factor categories, fine-grained health status analysis can be performed for a variety of environments and bridge section characteristics.

[0131] Step 132 counts the contribution of each category of influencing factors to the vibration of the bridge section based on the structure set obtained in step 131.

[0132] Example: The main vibration sources of a bridge section may be wind load (influence factor category 1) and vehicle load (influence factor category 2), which account for 60% and 40% of the vibration of the bridge section respectively; according to the load weight of the bridge section and the degree of vibration influence; the final result is a vector containing multiple influencing factor load coefficients, which describes the load characteristics of the current bridge section under different influencing factors; by calculating the load coefficient of the bridge section influencing factor, it is possible to quantify the extent to which different categories of influencing factors aggravate the vibration of the bridge section; calculating the load coefficient separately for each bridge section is helpful for hierarchical assessment of the importance and repair priority of the bridge section; as the load weight of the bridge section and the degree of vibration influence change, the influencing factor load coefficient can be dynamically updated to reflect the current actual working condition of the bridge.

[0133] Step 133 is based on the influence factor load coefficient of each bridge section, and the influence factor load coefficient of the whole bridge is obtained by summarizing and calculating the data of the whole bridge. By comprehensively considering the load coefficients of multiple bridge sections, the overall condition of the vibration load of the whole bridge is fully reflected; the influence factor load coefficient of the whole bridge can reflect the overall vibration load level of the bridge, and provide an important basis for the global health assessment of the bridge; through the summary analysis of the whole bridge data, the bridge section or influence factor that contributes the most to the overall health status can be found, and the allocation of maintenance resources for the whole bridge can be optimized; through the calculation of the influence factor load coefficient of the whole bridge, the overall monitoring of the health status of the bridge can be realized, and it is not limited to a single bridge section.

[0134] From the bridge section impact degree in step 131, the bridge section category in step 132 to the load coefficient of the whole bridge influencing factors in step 133, a multi-level health status assessment is achieved, which helps to fully understand the operating status of the bridge from the local to the overall; each step introduces a clear calculation formula to quantify the load effect of different influencing factors and improve the interpretability and operability of the assessment results; the entire process can be dynamically updated according to the real-time vibration data obtained, supporting the real-time and accuracy of bridge health monitoring; providing full-bridge and local load assessment results, providing data support for bridge maintenance decisions and reducing resource waste.

[0135] In a preferred embodiment of the present invention, the step 13 further comprises:

[0136] Step 134, judging the influence property of the vibration influence factor according to the load coefficient of the influence factor of the whole bridge; the influence property of the vibration influence factor includes direct influence and indirect influence;

[0137] Step 135, setting the impact factor load threshold, when the total bridge impact factor load coefficient is greater than the impact factor load threshold, then The attribute is directly affected; otherwise, The property is an indirect influence.

[0138] In the implementation of the present invention, step 134 divides the influencing properties of the vibration influencing factors into two categories:

[0139] Direct impact, the vibration influencing factor directly produces a significant load impact on the bridge (such as obvious structural damage or performance degradation) and indirect impact, the vibration influencing factor has a small load impact on the bridge or an indirect effect (such as indirect impact caused by other factors); according to the load coefficient of the entire bridge influencing factor calculated in step 133, combined with the load characteristics of the influencing factor classification, its specific impact nature is analyzed.

[0140] Direct influencing factors have a significant load-increasing effect on bridge vibration, which may directly lead to changes in the structural health status; indirect influencing factors have a relatively minor load effect on bridge vibration, which may be affected through the combined action of other factors; when a certain influencing factor (such as vehicle load) occupies a large proportion of the vibration load of the entire bridge, it may be a direct impact; while other factors (environmental changes, weather changes, aging of bridge section materials, and temperature changes) have a certain impact on vibration, but are more likely to affect bridge health indirectly by changing material properties; by distinguishing between direct and indirect impacts, clarifying the specific action mode of each influencing factor on bridge vibration will help to refine vibration management strategies; judging the nature of the impact will provide a causal basis for subsequent maintenance decisions, for example, direct influencing factors need to be intervened first, and indirect influencing factors can comprehensively analyze other action paths; further improving the accuracy of bridge vibration health assessment and providing more targeted support for full-bridge and local health management.

[0141] Step 135 sets an impact factor load threshold T according to the design specification or health assessment requirements of the bridge to distinguish between direct impact and indirect impact; the threshold can be flexibly adjusted according to different bridge structures and environmental conditions, for example, a lower threshold may be set for special environments (such as areas with strong winds or frequent earthquakes);

[0142] If the load coefficient of direct influence exceeds the threshold, it means that this factor plays a dominant role in bridge vibration and needs to be paid special attention to; if the load coefficient of indirect influence is lower than the threshold, it means that this factor has little effect on bridge vibration or affects it through other paths and is not the main contradiction.

[0143] Example analysis: For the influencing factors of a bridge vibration: the wind load vibration load coefficient F=0.8, if the threshold is set to 0.6, the wind load is judged to be a direct influence; the temperature influence load coefficient C=0.4, if it is lower than the threshold T=0.6, the temperature influence is judged to be an indirect influence; by setting the load threshold, the main (direct) vibration factors and the secondary (indirect) vibration factors can be quickly distinguished, so as to allocate maintenance resources more efficiently; the threshold can be flexibly adjusted to adapt to different bridge environmental conditions and design requirements; with the real-time update of vibration monitoring data, the threshold judgment result can also change dynamically to reflect the current bridge health status; direct influencing factors give priority to monitoring and repair (such as quickly reducing wind loads or traffic flow); indirect influencing factors can be comprehensively managed in combination with the overall bridge health status to reduce resource investment pressure.

[0144] In a preferred embodiment of the present invention, in step 13, the vibration deformation trend of the bridge section is obtained according to the influencing properties; based on the deformation trend, a bridge section stress tolerance balance strategy is obtained to balance the vibration stress tolerance of the bridge section, and further includes:

[0145] Step 136, based on the influence property of the vibration influence factor, obtain vibration response data; decompose the vibration signal mode according to the vibration response data to obtain multiple modal parameters;

[0146] Step 137, according to a plurality of modal parameters, by , to obtain the bridge segment deformation excitation vector; where, represents the bridge segment deformation excitation vector, K represents the stiffness matrix, C represents the damping matrix, M represents the mass matrix, and u(t) represents the displacement vector; according to the bridge segment deformation excitation vector, the vibration displacement change trend of the bridge segment is reflected, and the vibration displacement change trend is the vibration deformation trend of the bridge; when multiple modal parameters are within the preset threshold range, it indicates that the bridge segment deformation is stable or there is no deformation phenomenon;

[0147] Step 138, encoding the bridge segment deformation excitation vector into sequence data to train and construct a bridge segment dynamic deformation trend model; inputting at least one modal parameter in the bridge segment deformation excitation vector at the i-th time node into the bridge segment dynamic deformation trend model to output the bridge segment vibration deformation trend at the next time node;

[0148] Step 139, based on the output result of the dynamic deformation trend model of the bridge section, the internal stress distribution data of the bridge section is obtained through the established target optimization function and preset constraints; based on the internal stress distribution data of the bridge section, the stress bearing balance strategy of the bridge section is obtained.

[0149] In the implementation of the present invention, step 136 obtains relevant influencing factors through the influencing properties of the vibration influencing factors, and the influencing factors are the vibration response data of the bridge; and performs modal decomposition on them to understand the vibration characteristics of the bridge under external excitation; through the vibration response data of the bridge, the dynamic response of the bridge under various external forces is reflected; then, through the modal decomposition method (such as spectral analysis, time domain or frequency domain decomposition, etc.), the complex vibration signal is decomposed into multiple modes, each mode represents the vibration mode of the bridge at a specific frequency; important parameters such as the natural frequency, vibration mode and damping ratio of the bridge can be extracted, thereby providing a basis for the health status assessment of the bridge; through modal analysis, the dynamic characteristics of the bridge can be identified, including possible structural defects (such as local damage, fatigue cracks, etc.), providing direction for subsequent diagnosis and repair.

[0150] Step 137 constructs a deformation excitation vector of the bridge segment based on multiple modal parameters obtained by modal decomposition. This vector is composed of stiffness matrix (K), damping matrix (C), mass matrix (M) and displacement vector (u(t)) parameters; it describes the dynamic response of the bridge when subjected to external excitation, especially the deformation of each part of the bridge; thus, it can reveal the vibration displacement change trend of the bridge section, that is, the dynamic deformation mode of the bridge at different time points; the stiffness matrix (K), damping matrix (C), mass matrix (M) and displacement vector (u(t)) parameters have preset threshold ranges. Specifically, when the stiffness matrix parameter decreases by no more than 10%; the damping matrix parameter remains between 0.1% and 5%; the mass matrix parameter does not increase or decrease dramatically and the displacement vector parameter remains within 5mm to 20mm; the deformation of the bridge section is stable or there is no deformation phenomenon, the vibration response tends to be stable, indicating that the structure is healthy; otherwise, the vibration deformation trend is unstable, and there may be damage or abnormality; it helps to monitor the real-time dynamic response of the bridge, especially the deformation trend under the action of external loads; it can discover potential structural problems of the bridge in advance, such as stress concentration, crack expansion, etc., to ensure the safety of the bridge.

[0151] In step 138, the deformation excitation vector of the bridge section is encoded into sequence data so as to be input into a dynamic deformation trend model for training; the dynamic model can be a model based on machine learning (such as a deep learning model, a regression model, etc.), which is used to predict the dynamic deformation trend of the bridge in the future; by inputting modal parameters at each time node, the model can learn the law of the change of bridge vibration deformation over time, and then predict the future deformation trend; it can provide accurate prediction function for the long-term health monitoring of the bridge, and help foresee possible dynamic changes of the bridge in advance; through the prediction of dynamic deformation trend, maintenance decisions can be made for the bridge, such as arranging inspections or repairs in time when abnormal deformation trends are found, thereby extending the service life of the bridge.

[0152] Step 139 is based on the output result of the bridge section dynamic deformation trend model constructed in step 138, and uses the target optimization function and preset constraints to calculate to obtain the stress distribution inside the bridge section; the target optimization function is usually used to minimize or maximize a specific performance indicator (such as structural safety, minimum stress, etc.), and the constraints usually include material strength restrictions, geometric constraints, etc.; through simulation calculations, the stress distribution of the bridge section under specific working conditions can be obtained, so as to determine whether the bridge section is within the safety range and whether structural reinforcement or maintenance is required; by optimizing the stress distribution of the bridge section, it can be ensured that the stress borne by the bridge during operation is within a reasonable range to avoid safety problems such as overload and damage; optimizing the stress bearing balance strategy of the bridge section helps to reduce structural fatigue and damage and extend the service life of the bridge; it can be used as a basis for bridge maintenance and reinforcement, providing engineers with reliable data support to ensure the long-term safety of the bridge structure.

[0153] The working principles of these four steps are closely linked, and together they constitute a complete process for bridge health monitoring and intelligent management. Through the dynamic characteristics of the bridge, they provide a data basis for subsequent analysis; reveal the deformation trend of the bridge under dynamic loads to help discover potential structural problems; by predicting future deformation trends, they provide early warning of potential structural damage or instability; and by optimizing stress distribution, they ensure the safety and stability of the bridge under different working conditions.

[0154] In a preferred embodiment of the present invention, the bridge section stress bearing balance strategy in step 13 further includes:

[0155] According to the multiple modal parameters, deformation stress analysis is performed on the bridge segment to obtain bridge segment stress distribution data; based on the bridge segment stress distribution data, deformation of the bridge segment is graded and labeled to obtain graded labeled data; the graded labeled data includes a first-order deformation bridge segment, a second-order deformation bridge segment and a third-order deformation bridge segment;

[0156] In the implementation of the present invention, modal parameters refer to the characteristics of the bridge section structure under stress, vibration and other conditions, such as vibration frequency, vibration mode, damping ratio, etc. These modal parameters can be obtained through dynamic response testing of the bridge, or estimated through numerical simulation methods such as finite element analysis; through these modal parameters, the deformation of the bridge section under different vibration modes can be obtained, providing basic data for subsequent deformation stress analysis.

[0157] Based on the obtained modal parameters, deformation stress analysis is performed on the bridge section. This step is to calculate the stress distribution of the bridge under a specific load through structural mechanics analysis methods (such as finite element analysis, theoretical calculation, etc.). This analysis will take into account factors such as the bridge's geometry, material properties, and load conditions; through deformation stress analysis, stress distribution data of various parts of the bridge can be obtained, which reflects the possible deformation of the bridge under different working conditions.

[0158] The data obtained from stress analysis can be used to obtain the stress distribution of each bridge section in the bridge structure. These stress distribution data help determine which parts of the bridge section may be subjected to excessive stress, resulting in deformation or damage; based on the stress distribution data of the bridge section, deformation classification and labeling are performed. Different bridge sections are divided into different deformation grades according to their stress levels and deformation conditions. Common classification standards include:

[0159] like Figure 3 As shown, the first-order deformation bridge section has a small deformation, which is within the normal use range and may not have obvious damage.

[0160] like Figure 4 As shown in the figure, the deformation of the second-order deformation bridge section begins to appear, but is still within an acceptable range and may require regular monitoring and maintenance.

[0161] like Figure 5 As shown in the figure, the third-order deformation bridge section has serious deformation and may have cracks, deformation or other damage, and needs to be repaired or reinforced as soon as possible.

[0162] This hierarchical annotation can help engineers quickly identify possible risky parts in the bridge structure, thereby guiding subsequent maintenance and reinforcement work; the resulting hierarchical annotation data includes the deformation level of each bridge section, including first-order, second-order, and third-order deformation bridge sections. These data can be used for bridge health monitoring and subsequent decision-making, helping relevant personnel to carry out repairs and reinforcements in a timely manner to ensure the safe use of the bridge.

[0163] By conducting detailed deformation stress analysis and graded marking of bridges, potential structural problems can be discovered early, especially those with greater hidden dangers. By dealing with these problems in a timely manner, accidents can be avoided and the safety of bridges can be guaranteed; graded marking data makes bridge maintenance more accurate and scientific. According to the different deformation levels of bridge sections, relevant personnel can give priority to those bridge sections with larger deformations and higher risks, while for those with smaller deformations, the inspection cycle can be appropriately extended or low-cost preventive maintenance measures can be taken. This can ensure safety and reduce maintenance costs; regular stress analysis and graded marking can help detect early damage or fatigue of the structure, so as to carry out repairs and reinforcements as soon as possible. In this way, the aging and degradation process of the bridge can be effectively delayed and its service life can be extended.

[0164] By managing bridge sections at different levels, unnecessary comprehensive inspections and maintenance can be reduced. Only repairing bridge sections that require special attention can improve the efficiency of maintenance resources and reduce unnecessary maintenance expenses. By using graded annotation and stress analysis data, it can be combined with modern bridge health monitoring systems (such as sensor networks, big data analysis, etc.) to achieve intelligent management of bridges. This not only improves management efficiency, but also provides data support for long-term bridge operations.

[0165] In a preferred embodiment of the present invention, step 14 further comprises:

[0166] Step 141, based on the graded and labeled data, by distributing the load of the bridge so that the deformation stress of each bridge section tends to a stable state, so as to alleviate the stress fluctuation caused by vibration, and by strengthening the design to improve the coordinated bearing capacity between the bridge sections; by regularly obtaining the stress bearing health status of multiple graded and labeled bridge sections; according to the stress bearing health status of the bridge sections, the operation time and operation mechanism of the bridge sections are planned.

[0167] In the implementation of the present invention, step 141 is based on the load distribution of the graded annotation data, which provides the deformation levels of different bridge sections of the bridge (such as first-order, second-order, and third-order deformation bridge sections) and the corresponding stress bearing states. These data can accurately reflect the health status of the bridge section under different working conditions.

[0168] According to the graded and labeled data, the external loads borne by the bridge (such as vehicles, pedestrians, deadweight, wind force, etc.) are reasonably distributed to reduce the extra burden on high-stress bridge sections and avoid further damage. In terms of traffic organization, load limits are set, such as restricting the passage of heavy vehicles through severely damaged bridge sections, or directing traffic to bridge sections in better health. By adjusting the local structure of the bridge, such as adding auxiliary supports or reducing the deadweight of specific areas, deformation stress fluctuations can be alleviated. The deformation stress distribution of each bridge section is made more uniform to avoid local overload or stress concentration, and ultimately achieve a stable state of overall stress on the bridge.

[0169] Dynamic loads (such as impact force and vibration load caused by moving vehicles) will cause vibration of the bridge structure, and the stress will fluctuate over time. Stress fluctuations may accelerate fatigue damage, especially in bridge sections that have already deformed. Under dynamic conditions, traffic flow should be optimized (such as speed limits and lane adjustment) to reduce vibration amplitude. Vibration reduction devices (such as dampers) should be added between bridge sections to reduce vibration transmission. For bridge sections with large vibration amplitudes, local reinforcement design should be carried out, such as adding beams and support structures, to reduce local stress concentration. Stress fluctuations caused by vibration should be alleviated to delay bridge fatigue damage.

[0170] Collaborative load-bearing capacity refers to the ability of each bridge section to share the load. When some bridge sections are damaged, other bridge sections need to bear more load, which may cause further damage. According to the graded annotation data, the connection structure between the bridge sections is reinforced (such as strengthening the rigidity or flexibility of the connection parts); the redundancy of the structure is increased (such as setting up additional support structures) so that the load of the high-stress bridge section is partially transferred to the low-stress bridge section, thereby improving the overall collaborative load-bearing capacity; through the reinforcement design, the integrity and load-sharing capacity of the bridge are enhanced, thereby reducing the chain effect of local damage.

[0171] Stress bearing health status refers to the stress bearing capacity and health status of the bridge section under different working conditions. Through sensor networks, regular inspections and other means, the stress level and changes of the bridge section are monitored in real time or periodically; stress sensors, vibration sensors, etc. are installed to collect data on stress changes during bridge operation; the fatigue damage degree of the bridge section is analyzed using health monitoring algorithms to generate updated graded and labeled data; the health status of the bridge section is grasped in real time, potential problems are discovered in a timely manner, and data support is provided for operational decisions.

[0172] Develop a scientific operation schedule based on the stress-bearing health status of the bridge section. For example, for severely damaged bridge sections, heavy-load traffic can be restricted at night; or traffic can be reduced during peak hours to reduce the speed of damage; flexible traffic organization strategies can be formulated, such as setting up traffic diversion, speed reduction and speed limit, and closing heavy-load lanes; for some bridge sections that require long-term maintenance, the overall load-bearing mode or traffic mode of the bridge can be adjusted to avoid further damage to the structure; on the premise of ensuring safety, the service life of the bridge can be extended and the impact on traffic can be minimized; by optimizing load distribution and reinforcement design, the safety hazards caused by local stress concentration and vibration fluctuations of the bridge can be effectively reduced; the operation time and mechanism of the bridge section can be reasonably planned to avoid long-term overload of high-stress bridge sections and delay bridge fatigue and aging; based on the stress health status of the bridge section, accurate maintenance plans can be formulated to avoid indiscriminate comprehensive maintenance, thereby saving maintenance costs; by dynamically adjusting traffic organization and load distribution, the bridge's traffic capacity can be maintained to the maximum extent and traffic interruptions caused by structural damage can be reduced; with the help of health monitoring and graded labeling data, intelligent and scientific management of bridge operation can be realized, providing a basis for the management of the entire life cycle of the bridge; limited maintenance resources can be used first for high-risk bridge sections to improve maintenance efficiency and ensure the safety of key areas.

[0173] like Figure 2 As shown, a bridge vibration identification system based on artificial intelligence technology includes:

[0174] Data acquisition module: used to divide the bridge to be tested into multiple bridge sections, collect vibration data of the multiple bridge sections within a preset time, so as to obtain a bridge section vibration data set;

[0175] Data identification module: used for sequentially obtaining the vibration stress tolerance of multiple bridge sections, the vibration stress tolerance limits of multiple bridge sections, and the vibration influence degrees of multiple bridge sections based on the bridge section vibration data set;

[0176] Data processing module: it is used to obtain the vibration influence factor based on the vibration influence degree of the bridge section; to judge the influence property of the vibration influence factor based on the vibration influence factor; to obtain the vibration deformation trend of the bridge section according to the influence property; to obtain the bridge section stress tolerance balance strategy based on the deformation trend to balance the vibration stress tolerance of the bridge section;

[0177] Operation planning module: It is used to regularly obtain the stress tolerance health status of the bridge section based on the balanced bridge section vibration stress tolerance, so as to plan the operation time and operation mechanism of the bridge section.

[0178] When the functions of the above modules are implemented in the form of software functional units and used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks or optical disks.

[0179] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general intelligent device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or system. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

Claims

1. A bridge vibration identification method based on artificial intelligence technology, characterized in that: The following steps are involved: Dividing the bridge to be tested into a plurality of bridge sections, collecting vibration data of the plurality of bridge sections within a preset time period, so as to obtain a bridge section vibration data set; Based on the bridge section vibration data set, sequentially obtaining a plurality of bridge section vibration stress tolerances, a plurality of bridge section vibration stress tolerance limits, and a plurality of bridge section vibration influence degrees; The vibration data set of the bridge section collected within a preset time and a plurality of vibration stress tolerance limits of the bridge section are trained to construct a vibration recognition model, and the vibration recognition model outputs an identification result representing the vibration influence degree of the bridge section; at least one data item in the real-time monitored bridge section vibration data set is input into the vibration recognition model to output a prediction result of the vibration influence degree of the bridge section; The vibration recognition model comprises: Generate structural data from the bridge section vibration data set and a plurality of bridge section vibration stress tolerance limits within a preset time, and encode the structural data into sequence data to train and obtain the vibration recognition model; The sequence data is input into the vibration recognition model; the vibration recognition model includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output layer, and the intermediate representation data of multiple hidden layers are transmitted to the output layer, and the output layer output represents the recognition result of the vibration influence degree of the bridge section; Inputting at least one data item of the newly acquired bridge section vibration data set into the vibration recognition model, and outputting the prediction result of the bridge section vibration influence degree representing the newly acquired bridge section vibration data set at the output layer; Based on the vibration influence degree of the bridge section, a vibration influence factor is obtained; the vibration influence factor is used to judge the nature of the influence on the vibration influence degree of the bridge section.

2. The bridge vibration identification method based on artificial intelligence technology according to claim 1 is characterized in that: Based on the bridge section vibration data set, the vibration stress tolerance of multiple bridge sections is obtained, including: Preprocessing the bridge section vibration data set to obtain a preprocessed data set; extracting vibration features from the preprocessed data set to obtain a feature data set including time domain feature data and frequency domain feature data; According to the characteristic data set, the vibration stress change of the bridge section is obtained; based on the vibration stress change of the bridge section, the dynamic strain is obtained; the dynamic strain is converted into a vibration stress value; the vibration stress value is in a dynamically changing state, and the vibration stress value is converted into an equivalent stress scalar, and the equivalent stress scalar represents a comparable stress value; Based on the equivalent stress scalar, , to obtain the fatigue bearing capacity; the fatigue bearing capacity is the vibration stress bearing capacity of the bridge section, where, It represents the vibration stress tolerance of the bridge section, m represents the material coefficient of the bridge section, and K represents the fatigue constant.

3. The bridge vibration identification method based on artificial intelligence technology according to claim 2 is characterized in that: According to the vibration stress tolerance, multiple bridge section vibration stress tolerance limits are obtained, including: Based on the vibration stress tolerance, To obtain the fatigue limit strength of the bridge section; the fatigue limit strength of the bridge section is the vibration stress tolerance limit of the bridge section, where: Indicates the fatigue limit strength of the bridge segment; Based on the vibration stress tolerance limit of the bridge section, a phased assessment of the health status of the bridge section is performed; when the vibration stress tolerance limit of the bridge section >1, indicating that the bridge section is in a healthy state; when the vibration stress tolerance of the bridge section ≤1, indicating that the bridge segment has an imbalance risk; Based on the health status of the bridge section, the healthy state bridge section is checked to obtain the bridge section vibration stress tolerance correction limit; the bridge section vibration stress tolerance correction limit is used to reduce the error risk of the bridge section health status.

4. The bridge vibration identification method based on artificial intelligence technology according to claim 3 is characterized in that: According to the vibration stress tolerance limit of the bridge section, the vibration impact degree of multiple bridge sections is obtained, including: According to the health status of the bridge segment, ; To obtain the influence degree of bridge section vibration; The influence degree of bridge section vibration is used to measure the influence degree of vibration on the health status of the bridge section; Where, represents the influence degree of bridge section vibration, z represents the normalized index of influence degree, Indicates the modified limit of vibration stress tolerance of the bridge section; Based on the influence degree of the bridge section vibration, the influence results of the health status of the multiple bridge sections on the vibration are judged; the influence results include Indicates that the amplitude and frequency of the vibration signal are small and have no significant impact on the bridge structure; when Indicates that the vibration signal amplitude and frequency are close to the limit and need to be closely monitored; It means that the amplitude and frequency of the dynamic signal exceed the limit value, there is a risk of bridge section damage or fatigue, and repair measures need to be taken.

5. The bridge vibration identification method based on artificial intelligence technology according to claim 1 is characterized in that: Based on the vibration influencing factor, determining the influencing property of the vibration influencing factor; and obtaining the vibration deformation trend of the bridge section according to the influencing property; Based on the deformation trend, the bridge section stress balance strategy is obtained to balance the vibration stress tolerance of the bridge section; Based on the balanced bridge section vibration stress tolerance, the bridge section stress tolerance health status is obtained regularly; and according to the bridge section stress tolerance health status, the bridge section operation time and operation mechanism are planned.

6. The bridge vibration identification method based on artificial intelligence technology according to claim 5 is characterized in that: Based on the vibration influence degree of the bridge section, the vibration influence factor is obtained, including: The result of the influence of the vibration on the health status of the bridge section is obtained according to the influence degree of the vibration of the bridge section; multiple bridge sections are marked to obtain a structure set ; In the formula, n represents the impact factor category; According to the structure set, the bridge load weight is determined ,pass , to obtain the bridge section influence factor load coefficient; the bridge section influence factor load coefficient represents the degree of influence on the bridge section vibration aggravation; where, represents the load coefficient of the bridge section impact factor, Indicates the influence degree of bridge section vibration of the nth category influencing factor; Based on the load factor of the bridge section, To obtain the load coefficient of the whole bridge influence factor; set the load threshold of the influence factor, when the load coefficient of the whole bridge influence factor is greater than the impact factor load threshold, then The attribute is directly affected; otherwise, The property is an indirect influence.

7. The bridge vibration identification method based on artificial intelligence technology according to claim 6 is characterized in that: Obtaining the vibration deformation trend of the bridge section according to the influencing properties; Based on the deformation trend, the bridge section stress balance strategy is obtained to balance the vibration stress tolerance of the bridge section, including: Based on the influencing properties of the vibration influencing factors, vibration response data is obtained; according to the vibration response data, the vibration signal mode is decomposed to obtain a plurality of modal parameters; According to multiple modal parameters, a deformation excitation vector of the bridge section is obtained; according to the deformation excitation vector of the bridge section, the vibration displacement change trend of the bridge section is reflected, and the vibration displacement change trend is the vibration deformation trend of the bridge; when , it indicates that the deformation of the bridge section is stable or there is no deformation phenomenon; According to the bridge segment deformation excitation vector, the bridge segment deformation excitation vector is encoded into sequence data to train and construct a bridge segment dynamic deformation trend model; at least one modal parameter in the bridge segment deformation excitation vector at the i-th time node is input into the bridge segment dynamic deformation trend model to output the bridge segment vibration deformation trend at the next time node; Based on the output results of the dynamic deformation trend model of the bridge section, the internal stress distribution data of the bridge section and the stress balance strategy of the bridge section are obtained in turn through the established target optimization function and preset constraints.

8. The bridge vibration identification method based on artificial intelligence technology according to claim 7 is characterized in that: The bridge section stress bearing balance strategy includes: According to the plurality of modal parameters, deformation stress analysis is performed on the bridge section to obtain bridge section stress distribution data; based on the data, deformation of the bridge section is graded and labeled to obtain graded and labeled data including a first-order deformation bridge section, a second-order deformation bridge section and a third-order deformation bridge section; Based on the graded and labeled data, the load of the bridge is distributed so that the deformation stress of each bridge section tends to a stable state to alleviate the stress fluctuation caused by vibration, and the coordinated bearing capacity between the bridge sections is improved through reinforcement design; by regularly obtaining the stress bearing health status of multiple graded and labeled bridge sections, the operation time and operation mechanism of the bridge sections are planned.

9. A bridge vibration identification system based on artificial intelligence technology, characterized in that: The system is provided with an electronic device including a memory, a processor, and a bridge vibration identification method program based on artificial intelligence technology stored in the memory and executable on the processor, wherein the program is used to implement the bridge vibration identification method according to any one of claims 1 to 8 when executed by the processor, and the system includes: Data acquisition module: used to divide the bridge to be tested into multiple bridge sections, collect vibration data of the multiple bridge sections within a preset time, so as to obtain a bridge section vibration data set; Data identification module: used for sequentially obtaining the vibration stress tolerance of multiple bridge sections, the vibration stress tolerance limits of multiple bridge sections, and the vibration influence degrees of multiple bridge sections based on the bridge section vibration data set; Data processing module: it is used to obtain the vibration influence factor based on the vibration influence degree of the bridge section; to judge the influence property of the vibration influence factor based on the vibration influence factor; to obtain the vibration deformation trend of the bridge section according to the influence property; to obtain the bridge section stress tolerance balance strategy based on the deformation trend to balance the vibration stress tolerance of the bridge section; Operation planning module: It is used to regularly obtain the stress tolerance health status of the bridge section based on the balanced bridge section vibration stress tolerance, so as to plan the operation time and operation mechanism of the bridge section.

Images