Post-earthquake structure evaluation and restoration method and device based on two-stage time history analysis

Abstract

The invention provides a post-earthquake structure evaluation and repair method and device based on two-stage time-history analysis, and relates to the technical field of structure damage evaluation. The method comprises the following steps: performing two-stage elastic-plastic time-history analysis on a multi-layer structure through a fiber beam model to obtain structure response data; building a deep learning model based on MLP and LSTM, and predicting structure response data of a second stage; performing weighted summation dimension reduction processing on the predicted structure response data of the second stage through an evaluation index dimension reduction method to obtain a one-dimensional damage evaluation index after the earthquake of the structure, and calculating 95% quantile of a probability density curve; constructing a one-dimensional damage evaluation index limit value of the structural mechanical response state as a damage grade division basis; according to the post-earthquake structure safety assessment method and the post-earthquake structure safety assessment system, rapid assessment of post-earthquake structure safety can be achieved, and a basis is provided for post-earthquake repair decision making.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of structural damage assessment, in particular to a post-earthquake structure assessment and repair method and device based on two-stage time history analysis. BACKGROUND

[0002] Earthquake disasters can have a serious impact on the seismic performance of building structures. In recent years, a large number of studies have focused on pre-earthquake seismic design, effectively enhancing the seismic resilience and disaster tolerance of building structures. In this context, to further reduce the losses caused by earthquakes, research on building structures in the epicenter and post-earthquake stages is still needed. The destruction of earthquake disasters to building structures has a cumulative effect. After the initial earthquake, structures that have not completely collapsed may experience irreversible effects such as stiffness degradation and cumulative plastic damage. Efficiently assessing the safety performance of structures after an earthquake and taking timely and effective measures can reduce losses in terms of manpower, material resources, and financial resources.

[0003] Traditional post-earthquake assessment relies on manual on-site detection, which divides damage levels by observing surface features such as crack distribution and deformation degree, but has the defects of low efficiency, strong subjectivity, and high risk. With the development of intelligent monitoring technology, damage identification methods based on structural dynamic response parameters (such as natural frequency, inter-story drift angle, and plastic hinge distribution) have gradually become a research hotspot. For example, Cawley et al. achieved two-dimensional structure damage detection by relating the frequency change ratio to stiffness degradation; Yang Yongqiang et al. proposed a combination of maximum inter-story drift angle and natural frequency to evaluate the damage index. These methods achieve damage quantification through non-contact monitoring data, but are still limited by the lack of parameter sensitivity and the reliance on empirical formulas.

[0004] In the 21st century, breakthroughs in machine learning technology have provided new solutions for complex nonlinear problems. Some researchers have trained models to identify damage states using structural displacement response data, while others have used BP neural networks to analyze the mapping relationship between reinforced concrete structure parameters and damage, demonstrating the engineering applicability of machine learning methods. These studies have promoted the paradigm shift from empirical formulas to data-driven damage assessment, but still have two common problems: the evaluation indicators used are difficult to obtain, such as cracks and local buckling (which require carpet-like screening into the structure) or structural fundamental frequencies (existing detection methods are difficult to excite large structures); ideal structures are used as research objects, without considering the impact of post-earthquake damage states on subsequent mechanical responses of structures.

[0005] At present, there are few secondary damage evaluation methods based on deep learning combined framework structure, and the existing related methods often evaluate damage based on the characteristics of ideal structures, only consider the response (such as interlayer displacement, crack) and static characteristics such as design parameters generated after the structure first experiences an earthquake as the basis to reflect the damage degree, and rarely have research on the response of the damaged structure after experiencing an earthquake again to characterize its safety performance.

[0006] Traditional post-earthquake evaluation methods mainly rely on artificial field detection, and the damage level is divided by observing the apparent characteristics such as crack distribution and deformation degree, but there are defects such as low efficiency, strong subjectivity and poor real-time performance. Although the damage identification method based on dynamic response parameters (such as frequency change ratio and interlayer displacement angle) realizes non-contact monitoring, it is limited by the insufficient parameter sensitivity and the dependence on empirical formula. The introduction of machine learning technology promotes the data-driven transformation of damage evaluation, for example, Stephens et al. use neural networks to predict post-earthquake damage state, and Arslan et al. analyze the mapping relationship between steel reinforced concrete structure parameters and damage through BP neural network, but the evaluation indicators based on the existing research are difficult to obtain, such as cracks, local buckling (need to enter the structure for carpet screening), or structure basic frequency (existing detection means is difficult to excite large structures), and the ideal structure is taken as the research object, without considering the influence of post-damage state on subsequent mechanical response of the structure, and there is a lack of research on safety performance evaluation considering the response of the damaged structure after experiencing an earthquake again. SUMMARY

[0007] In order to solve the problem that the existing method is inefficient, dangerous and easily affected by subjective experience according to the on-site detection and evaluation of experienced experts; the data-driven damage evaluation method realizes efficient structure evaluation by analyzing the mapping relationship between structure parameters and damage, but the parameter is single, and generally only considers the state of ideal structure as input, and the change of actual post-earthquake structure state is less considered. At present, there is a lack of evaluation method considering the response of the damaged structure after experiencing an earthquake again during the subsequent service process, and the evaluation of the damaged structure is not comprehensive and reliable enough, and the embodiment of the present application provides a post-earthquake structure evaluation and repair method and device based on two-stage time history analysis, which no longer considers only the structure state under ideal state, and predicts the damage state probability density distribution of the post-earthquake structure after experiencing an earthquake again by combining a large amount of time series dynamic seismic wave data with the response index data of the post-earthquake structure, to realize fast and accurate damage evaluation and repair decision. The technical solution is as follows:

[0008] On the one hand, a post-earthquake structure evaluation and repair method based on two-stage time history analysis is provided, and the method is realized by a post-earthquake structure evaluation and repair device, and the method comprises:

[0009] Obtain the structure attribute parameter data and the structure response data of the post-earthquake structure to be evaluated, input the structure attribute parameter data and the structure response data of the post-earthquake structure to be evaluated into the constructed evaluation model, and obtain the safety evaluation result and the repair decision of the post-earthquake structure.

[0010] The construction process of the evaluation model comprises:

[0011] S1, based on the structure attribute parameter data, a multi-layer structure is constructed, based on the collected multiple seismic wave acceleration sequence data, the multi-layer structure is subjected to multiple two-stage elastic-plastic time-history analyses through a fiber beam model, various working conditions of the multi-layer structure subjected to two earthquakes are simulated, structure response data of the first stage and corresponding structure response data of the second stage under various two-stage working conditions of the multi-layer structure are obtained, and a training data set is constructed according to the structure attribute parameter data, the structure response data of the first stage, the second-stage seismic wave acceleration sequence data and the structure response data of the second stage.

[0012] S2, a deep learning model is built based on a multi-layer perception MLP and a long short-term memory neural network LSTM, the deep learning model is trained according to the training data set, a trained deep learning model is obtained, and the trained deep learning model has the ability to predict the structure response data of the second stage by inputting the structure attribute parameter data, the structure response data of the first stage and the second-stage seismic wave acceleration sequence data.

[0013] S3, the maximum interlayer displacement angle data of each layer or the plastic hinge rate data of each layer in the predicted structure response data of the second stage are subjected to weighted sum dimension reduction processing through an evaluation index dimension reduction method, one-dimensional damage evaluation index data of the structure after experiencing the second-stage seismic wave is obtained, a probability density curve is constructed according to a data set composed of respective one-dimensional damage evaluation indexes of the structure after experiencing various second-stage seismic waves, and a 95% quantile of the probability density curve is obtained.

[0014] S4, a damage state is defined, a critical displacement angle between damage states is calculated, a structure mechanics response state corresponding to the critical displacement angle between damage states is obtained, and the structure mechanics response state is represented by response data; the response data is subjected to weighted sum dimension reduction processing through an evaluation index dimension reduction method, and a one-dimensional damage evaluation index limit value of the structure mechanics response state is obtained as a damage grade division basis.

[0015] S5, according to the 95% quantile of the probability density curve and the damage grade division basis, a post-earthquake frame structure safety evaluation result and a repair decision are obtained.

[0016] Optionally, based on the collected multiple seismic wave acceleration sequence data, the multiple two-stage elastic-plastic time-history analysis of the multi-layer structure is performed through the fiber beam model in S1, various working conditions of the multi-layer structure are simulated after experiencing two earthquakes in sequence, and the first-stage structural response data and the corresponding second-stage structural response data of the multi-layer structure under various two-stage working conditions are obtained, including:

[0017] Based on the first-stage seismic wave acceleration sequence data, the first-stage elastic-plastic time-history analysis of the multi-layer structure is performed through the fiber beam model, and the first-stage structural response data is obtained; wherein the first-stage structural response data includes the residual inter-story drift angle data or the plastic hinge rate data of each layer of the first-stage damaged structure.

[0018] Based on the second-stage seismic wave acceleration sequence data, the second-stage elastic-plastic time-history analysis of the multi-layer structure after the first-stage earthquake is performed through the fiber beam model, and the second-stage structural response data is obtained; wherein the second-stage structural response data includes the maximum inter-story drift angle data or the plastic hinge rate data of each layer of the second-stage damaged structure.

[0019] Optionally, the second-stage structural response data is predicted by inputting the structural attribute parameter data, the first-stage structural response data and the second-stage seismic wave acceleration sequence data in S2, including:

[0020] The structural attribute parameter data and the first-stage structural response data are input into the multi-layer perception machine MLP, and the multi-layer perception machine MLP output is obtained through three fully connected layers.

[0021] The second-stage seismic wave acceleration sequence data is input into the long short-term memory neural network LSTM, and the long short-term memory neural network LSTM output is obtained.

[0022] The multi-layer perception machine MLP output and the long short-term memory neural network LSTM output are input into three fully connected layers, and the second-stage structural response data is predicted.

[0023] Optionally, the evaluation index dimension reduction method in S3 includes:

[0024] The maximum inter-story drift angle data or the plastic hinge rate data of each layer in the predicted second-stage structural response data is weighted and summed according to a preset proportion to obtain one-dimensional damage evaluation index data of the structure after experiencing the second-stage seismic wave.

[0025] The calculation formula of the evaluation index dimension reduction method is shown in the following formula (1):

[0026] (1)

[0027] In the formula, This represents a one-dimensional damage assessment index of the structure after experiencing the second phase of the seismic wave. Represents the post-earthquake response matrix of the structure. Indicates matrix transpose. This represents the weight matrix.

[0028] Optionally, the damage state defined in S4 includes:

[0029] By performing lateral pushover analysis on a multi-layer structure, the load-displacement angle curve of the multi-layer structure under pushover action is obtained, the peak point and limit point of the load-displacement angle curve are determined, the yield point of the load-displacement angle curve is determined by the farthest point method, and then the damage state is defined; among them, the damage state is used to describe the fully elastic, yield, peak and limit states of the multi-layer structure.

[0030] Optionally, the calculation of the critical displacement angle between different damage states in S4 includes:

[0031] The critical displacement angle between the first-level damage states is calculated as shown in equation (2) below:

[0032] (2)

[0033] The critical displacement angle between the second-level damage states is calculated as shown in equation (3):

[0034] (3)

[0035] The critical displacement angle between the third-level damage states is calculated as shown in equation (4):

[0036] (4)

[0037] In the formula, Indicates the critical displacement angle 1. This represents the displacement angle corresponding to the yield point. Indicates the critical displacement angle 2. This represents the displacement angle corresponding to the peak point. Indicates the critical displacement angle 3. This represents the displacement angle corresponding to the extreme point.

[0038] Optionally, S5, based on the 95th quantile of the probability density curve and the damage level classification criteria, yields the post-earthquake frame structure safety assessment results and repair decisions, including:

[0039] When the 95th percentile of the probability density curve is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, the damage level is determined to be no damage, and the person can move in immediately.

[0040] When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, the damage level is determined to be minor damage, requiring minor repairs before relocation.

[0041] When the 95th quantile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be moderate damage, and repairs are required before occupancy.

[0042] When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be severe damage, and the structure needs to be rebuilt.

[0043] On the other hand, a post-earthquake structural assessment and repair device based on two-stage time history analysis is provided. This device is applied to a post-earthquake structural assessment and repair method based on two-stage time history analysis. The device includes:

[0044] The assessment module is used to acquire structural attribute parameter data and structural response data of the post-earthquake structure to be assessed. The structural attribute parameter data and structural response data of the post-earthquake structure to be assessed are input into the constructed assessment model to obtain the safety assessment results and repair decisions of the post-earthquake structure.

[0045] The process of constructing the evaluation model includes:

[0046] S1. Construct a multi-story structure based on structural property parameter data. Based on the collected seismic wave acceleration sequence data, perform multiple two-stage elastoplastic time history analyses on the multi-story structure using a fiber beam model. Simulate various working conditions of the multi-story structure under two earthquakes. Obtain the structural response data of the first stage and the corresponding structural response data of the second stage under various two-stage working conditions of the multi-story structure. Construct a training dataset based on the structural property parameter data, the structural response data of the first stage, the seismic wave acceleration sequence data of the second stage, and the structural response data of the second stage.

[0047] S2. A deep learning model is built based on a multilayer perceptron (MLP) and a long short-term memory (LSTM) neural network. The deep learning model is trained using the training dataset to obtain a trained deep learning model. The trained deep learning model has the ability to predict the structural response data of the second stage by inputting structural attribute parameter data, structural response data of the first stage, and seismic wave acceleration sequence data of the second stage.

[0048] S3. The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed to reduce the dimension of the data by using the evaluation index dimensionality reduction method. This yields one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic waves. Based on the dataset composed of the one-dimensional damage assessment indexes of the structure after experiencing various second-stage seismic waves, a probability density curve is constructed, and the 95th percentile of the probability density curve is obtained.

[0049] S4. Define the damage state, calculate the critical displacement angle between each level of damage state, and obtain the structural mechanical response state corresponding to the critical displacement angle between each level of damage state. The structural mechanical response state is characterized by the response data. The response data is weighted and summed to reduce the dimensionality of the evaluation index, and the one-dimensional damage evaluation index limit of the structural mechanical response state is obtained as the basis for damage level classification.

[0050] S5. Based on the 95th percentile of the probability density curve and the damage level classification criteria, the post-earthquake frame structure safety assessment results and repair decisions are obtained.

[0051] On the other hand, a post-earthquake structural assessment and repair device is provided, the post-earthquake structural assessment and repair device comprising: a processor; a memory, the memory storing computer-readable instructions, which, when executed by the processor, implement any of the methods described above for post-earthquake structural assessment and repair based on two-stage time history analysis.

[0052] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, the at least one instruction being loaded and executed by a processor to implement any of the above-described methods for post-earthquake structural assessment and repair based on two-stage time history analysis.

[0053] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0054] This invention addresses the need for rapid and accurate seismic performance assessment of existing composite frame structures by proposing a deep learning-based method for post-earthquake safety assessment of frame structures. During the post-earthquake assessment phase, the method considers the responses characteristic of the structure to subsequent seismic waves of different waveforms and intensities. Based on these responses, the current damage status of the structure is assessed, and repair decisions are made. This improves the efficiency and reliability of subsequent seismic performance assessments of damaged structures, promoting further development and application in the field of structural damage assessment. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is a flowchart of a post-earthquake structural assessment and repair method based on two-stage time history analysis provided by an embodiment of the present invention;

[0057] Figure 2 This is a comparison chart of actual values ​​and predicted values ​​provided in an embodiment of the present invention;

[0058] Figure 3 This is a model architecture diagram provided in an embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram of the damage assessment index dimensionality reduction method provided in the embodiments of the present invention;

[0060] Figure 5 This is a schematic diagram of the process for obtaining the limit value of the one-dimensional damage assessment index provided in an embodiment of the present invention;

[0061] Figure 6 This is a schematic diagram of the 95th percentile of the probability density distribution of the evaluation index provided in the embodiments of the present invention;

[0062] Figure 7 This is a schematic diagram of the evaluation process provided in an embodiment of the present invention;

[0063] Figure 8 This is a block diagram of a post-earthquake structural assessment and repair device based on two-stage time history analysis provided in an embodiment of the present invention;

[0064] Figure 9 This is a schematic diagram of a post-earthquake structural assessment and repair device provided in an embodiment of the present invention. Detailed Implementation

[0065] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0066] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0067] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0068] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0069] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0070] This invention provides a post-earthquake structural assessment and repair method based on two-stage time history analysis. This method can be implemented using post-earthquake structural assessment and repair equipment, which can be a terminal or a server. Figure 1 The flowchart shown is for a post-earthquake structural assessment and repair method based on two-stage time history analysis. The process of this method may include the following steps:

[0071] S1. Construct a multi-story structure based on structural property parameter data. Based on the collected seismic wave acceleration sequence data, perform a large number of two-stage elastoplastic time history analyses on the multi-story structure using a fiber beam model. Simulate a large number of working conditions of the multi-story structure experiencing two earthquakes. Obtain the first-stage structural response data and the corresponding second-stage structural response data of the multi-story structure under various two-stage working conditions. Construct a training dataset based on the structural property parameter data, the first-stage structural response data, the second-stage seismic wave acceleration sequence data, and the second-stage structural response data.

[0072] In one feasible implementation, structural attribute parameter data may include: number of stories; span of the frame structure in the X and Y directions; number of spans in the X and Y directions; number of floors; height of the standard floor; height of the ground floor; column section type (e.g., rectangular = 0, I-shaped = 1, composite section = 2); column section dimensions; column section material strength; column section reinforcement ratio (for steel structures, this can represent stiffening rib density; for composite structures, it represents the arrangement density of connectors such as studs); beam section type; beam section dimensions; beam section material strength; beam section reinforcement ratio (same as above, changing parameter types according to different structures); seismic intensity; characteristic period; damping ratio. These parameters may have different data types for different structures, but they can all be addressed through coding.

[0073] This invention can evaluate various types of structures, including multi-story structures such as reinforced concrete frame structures, steel frame structures, and composite frame structures.

[0074] Optionally, in step S1 above, based on the collected multiple seismic wave acceleration sequence data, multiple two-stage elastoplastic time history analyses are performed on the multi-story structure using the finite element software MSC.MARC to simulate various working conditions of the multi-story structure undergoing two earthquakes, obtaining the first-stage structural response data and the corresponding second-stage structural response data under various two-stage working conditions of the multi-story structure. This may include:

[0075] Based on the first-stage seismic wave acceleration sequence data, a first-stage elastoplastic time history analysis was conducted on a multi-story structure using a fiber beam model to obtain the first-stage structural response data. The first-stage structural response data includes: residual inter-story drift angle data or plastic hinge rate data of each story in the first-stage damaged structure.

[0076] Based on the second-stage seismic wave acceleration sequence data, a second-stage elastoplastic time history analysis was conducted on the multi-story structure after the first-stage earthquake using a fiber beam model to obtain the second-stage structural response data. The second-stage structural response data includes: the maximum inter-story drift angle data or the plastic hinge ratio data of each story of the earthquake-damaged structure in the second stage.

[0077] In one feasible implementation, a two-stage elastoplastic time history analysis is conducted using a fiber beam model to construct a dataset. The response of the structure after experiencing the second earthquake is calculated using the initial post-earthquake state and secondary seismic wave data of the composite frame structure, thus constructing a post-earthquake structural dynamic response database. This invention utilizes a fiber beam model developed based on the MSC.MARC general finite element analysis program to conduct multiple earthquake simulations on a ten-story steel-concrete composite frame structure. Two stages of seismic loads are applied to the prototype structure: In the first stage, fourteen different seismic waves with PGA (Peak Ground Acceleration) values ​​of 0.20g, 0.30g, and 0.40g are input to the structural base to induce different degrees of damage, generating a total of 42 load cases. In the second stage, fourteen seismic waves with different PGA values ​​(0.20g, 0.30g, and 0.40g) are used to perform elastoplastic time history analysis on the damaged structural model formed in the first stage, resulting in 42 additional load cases arising from the damage results of each load case in the first stage. The two phases generated a total of 42 × 42 = 1764 possible operating conditions. Simultaneously, a 3-second time interval was set between the two earthquake phases to ensure that the post-earthquake structure could recover to a static state before the second phase earthquake. Finally, the structural response data for each phase were extracted to form a dataset.

[0078] S2. A deep learning model is built based on a multilayer perceptron (MLP) and a long short-term memory (LSTM) neural network. The deep learning model is trained using the training dataset to obtain a trained deep learning model. The trained deep learning model has the ability to predict the structural response data of the second stage by inputting structural attribute parameter data, structural response data of the first stage, and seismic wave acceleration sequence data of the second stage.

[0079] Optionally, in S2, by inputting structural attribute parameter data, first-stage structural response data, and second-stage seismic wave acceleration sequence data, the second-stage structural response data is predicted, including:

[0080] Structural property parameter data and structural response data for the first stage under various working conditions are input into a multilayer perceptron (MLP), and the output of the MLP is obtained after passing through three fully connected layers.

[0081] The second-stage seismic wave acceleration sequence data under various working conditions are input into the Long Short-Term Memory Neural Network (LSTM) to obtain the output of the LSTM.

[0082] By passing the outputs of the Multilayer Perceptron (MLP) and the Long Short-Term Memory Neural Network (LSTM) through three fully connected layers, the predicted structural response data for the second stage after experiencing multiple different seismic waves is obtained.

[0083] In one feasible implementation, a deep learning model is constructed and trained to achieve predictive performance. A deep learning model is built by combining MLP (Multilayer Perceptron) and LSTM (Long Short-Term Memory). The deep learning model takes residual inter-story drift angle data or plastic hinge rate data of each layer of the post-earthquake structure and seismic wave acceleration sequence data from the second-stage earthquake as input. First, the MLP receives structural attribute parameter data and residual inter-story drift angle data or plastic hinge rate data of each layer of the damaged structure, and outputs 1 through three fully connected layers. The LSTM receives the second seismic wave sequence data and outputs 2. Outputs 1 and 2 are concatenated and used as input again through three fully connected layers. Finally, the mechanical response of the damaged structure under the action of the second-stage seismic wave is output, achieving rapid and accurate prediction of the mechanical response of the post-earthquake structure under subsequent earthquake action based on the post-earthquake structural damage state and secondary seismic wave information. The prediction performance of the deep learning model is as follows: Figure 2 As shown, the deep learning model architecture is as follows Figure 3 As shown.

[0084] This invention combines a composite feature model of MLP and LSTM: by combining MLP and LSTM to build a deep learning model for the problem of this invention, the model is input with the response of the damaged composite frame structure and secondary seismic wave data, performs composite feature prediction, and outputs the secondary post-earthquake response of the damaged structure. Compared with other methods for obtaining the secondary post-earthquake response of structures, the efficiency is greatly improved.

[0085] S3. The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed to reduce the dimension of the data by using the evaluation index dimensionality reduction method. This yields one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic waves. Based on the dataset composed of the one-dimensional damage assessment indexes of the structure after experiencing various second-stage seismic waves, a probability density curve is constructed, and the 95th percentile of the probability density curve is obtained.

[0086] In one feasible implementation, a one-dimensional damage assessment index for the frame structure is obtained through deep learning model prediction and weight allocation calculation. The result predicted by the deep learning model is multi-dimensional post-earthquake response data of the structure, such as the residual displacement angle of each floor of a ten-story structure. To concisely and efficiently characterize the structural damage state, this invention proposes a dimensionality reduction method for the assessment index, as follows:

[0087] The model predicts the post-earthquake structural response state, such as the residual inter-story drift angle and plastic hinge ratio of each of the ten stories. The structural response data of each story are weighted and summed according to a certain proportion to obtain the one-dimensional damage assessment index of the response state. The calculation method is shown in formula (1):

[0088] (1)

[0089] In the formula, It serves as a one-dimensional damage assessment index for structures after an earthquake. The post-earthquake response matrix of the structure can be composed of inter-story drift angles or plastic hinge ratios, etc. This is the weight matrix, composed of the weight coefficients of each layer of the structure. The calculation method is as follows: Figure 4 As shown.

[0090] This invention provides an assessment process based on post-earthquake response data of a structure after a secondary earthquake: the model predicts the response data of the damaged structure after experiencing different secondary earthquakes, and then, based on the damage level classification criteria derived from pushover analysis, the probability density curve of the one-dimensional damage assessment index of the damaged structure after the secondary earthquake is obtained. The 95th percentile of the probability density curve is calculated to determine the safety of the structure in the subsequent service stage and to make a repair decision.

[0091] S4. Define the damage state, calculate the critical displacement angle between each level of damage state, and obtain the structural mechanical response state corresponding to the critical displacement angle between each level of damage state. The structural mechanical response state is characterized by the response data. The response data is weighted and summed to reduce the dimensionality of the evaluation index, and the one-dimensional damage evaluation index limit of the structural mechanical response state is obtained as the basis for damage level classification.

[0092] In one feasible implementation, damage level classification is obtained through pushover analysis and weight allocation. Lateral pushover analysis of the composite frame structure yields the load-displacement angle curve under pushover action, and the peak point of the curve is determined. and the limit point The yield point of the curve is determined using the farthest point method. ,in, This represents the peak displacement corresponding to the peak point. Based on this, damage states DS0~DS3 are defined to describe the fully elastic, yielding, peak, and ultimate states of the structure. The critical displacement angle between each damage state is... The calculation method is shown in formulas (2)-(4). The critical displacement angle between each damage state is calculated, thereby obtaining the structural mechanical response state corresponding to each critical displacement angle. The structural mechanical response state is characterized by a large amount of response data. The multidimensional response data is weighted and summed according to the method in step S3 to reduce the dimension, and the one-dimensional damage assessment index limit of the structural mechanical response state corresponding to each critical displacement angle is obtained. , and This serves as the basis for classifying damage levels. The process for obtaining the limits of one-dimensional damage assessment indicators is as follows: Figure 5 As shown:

[0093] (2)

[0094] The critical displacement angle between the second-level damage states is calculated as shown in equation (3):

[0095] (3)

[0096] The critical displacement angle between the third-level damage states is calculated as shown in equation (4):

[0097] (4)

[0098] In the formula, Indicates the critical displacement angle 1. This represents the displacement angle corresponding to the yield point. Indicates the critical displacement angle 2. This represents the displacement angle corresponding to the peak point. Indicates the critical displacement angle 3. This represents the displacement angle corresponding to the extreme point.

[0099] S5. Based on the 95th percentile of the probability density curve and the damage level classification criteria, the post-earthquake frame structure safety assessment results and repair decisions are obtained.

[0100] Specifically, when the 95th percentile of the probability density curve is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, the damage level is determined to be no damage, and the person can move in immediately.

[0101] When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, the damage level is determined to be minor damage, requiring minor repairs before relocation.

[0102] When the 95th quantile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be moderate damage, and repairs are required before occupancy.

[0103] When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be severe damage, and the structure needs to be rebuilt.

[0104] In one feasible implementation, a post-earthquake safety performance assessment and repair decision-making process for composite frame structures is proposed based on the prediction results of a deep learning model. After the structure experiences its first earthquake, structural damage state data (residual inter-story displacement and plastic hinge ratio) are obtained through detection methods. This damage state data is combined with a large amount of seismic wave acceleration sequence data as input to a pre-trained model. This model predicts the response datasets of the structure after different types of secondary earthquakes, referred to as the re-earthquake response dataset. Each type of re-earthquake response data is used to obtain its own one-dimensional damage assessment index according to the method given in step S3. This allows us to obtain the probability density curve of the one-dimensional damage assessment index for the re-earthquake response dataset, and calculate the 95th quantile of the probability density of the damage assessment index. ,like Figure 6 As shown, This means that the area to the left of the value (less than or equal to the value) accounts for 95% of the total area. The calculation formula is shown in formula (5) (taking a general normal distribution as an example). The mean of the distribution. Let be the standard deviation of the distribution. According to the proposed criteria for classifying damage levels, when At that time, the damage level is DS0 (no damage), and the person can move in immediately; when At that time, the damage level was DS1 (minor damage), requiring minor repairs before occupancy; when At that time, the damage level was DS2 (moderate damage), requiring repair before occupancy; when At that time, the damage level was DS3 (severe damage), and the structure required reconstruction. The assessment and decision-making process is as follows: Figure 7 As shown:

[0105] (5)

[0106] By employing the aforementioned technical solutions and methods, and fully considering the cumulative damage from the elastoplastic response of the composite frame structure and the influence of different secondary seismic waves, the secondary damage response of the structure is predicted. Based on the secondary response results, an efficient and accurate safety performance assessment of the post-earthquake structure is achieved, enabling rapid and reliable repair decisions.

[0107] Obtain structural attribute parameter data and structural response data of the post-earthquake structure to be evaluated, input the structural attribute parameter data and structural response data of the post-earthquake structure to be evaluated into the constructed evaluation model, and obtain the safety assessment results and repair decisions of the post-earthquake structure.

[0108] In one feasible implementation, the specific usage of the present invention is as follows: after a structure experiences an earthquake, staff collect the structure's property parameter data and the post-earthquake mechanical response data (residual displacement angle and plastic hinge ratio). The parameter data and mechanical response data are then combined and input into the program launched by the present invention (which includes a deep learning prediction model and an evaluation program). The program will quickly assess the damage level of the structure and provide repair decision suggestions.

[0109] This invention utilizes acquired initial post-earthquake damage data of composite frame structures, such as residual inter-story drift angles and plastic hinge ratios, to predict secondary post-earthquake response data (such as maximum inter-story drift angles and plastic hinge ratios) based on a trained composite deep learning model. From the predicted secondary response data, a one-dimensional damage assessment index is calculated using the method proposed in this paper, and the 95th percentile of the probability density curve of this one-dimensional damage assessment index is calculated. ,according to The method uses damage level classification based on pushover analysis to determine the damage level and provide a repair decision plan. This method can achieve rapid assessment of post-earthquake structural safety and provide a basis for post-earthquake repair decisions.

[0110] Patents CN116467789A, CN117251926A, and CN119962304A are mainly based on the recalculation of seismic waves for evaluation. Their input parameters (such as the distance between the seismic record point and the structure, site soil conditions, and non-structural component load conditions) do not include the actual state characteristics of the structure after the earthquake, resulting in a high degree of distortion in the evaluation results.

[0111] While patent CN113435091A considers the construction dates of buildings and bridges and the required data is relatively easy to obtain, it still fails to incorporate current structural damage data into the model, resulting in distorted evaluation results. In contrast, the structural parameter data required by the proposed solution is also readily available on a macroscopic scale, and further incorporates structural damage information by using response data after the initial earthquake (such as residual displacement angle and plastic hinge ratio) as input.

[0112] Patent CN113868750A corrects the degradation parameters of structural mechanical properties based on on-site damage photographs. Its degradation prediction relies on a large accumulation of experimental data, and its advantage lies in utilizing on-site feedback information, relatively reducing distortion. However, the required experimental damage image data is difficult to obtain sufficiently, and there are differences between experimental component images and actual post-renovation damage images, affecting the accuracy of the input data and placing high demands on the input data. Furthermore, the degradation indices predicted by this method need to be used to adjust the parameters of the structural mechanical model, followed by finite element calculations, which is time-consuming. This patent uses a static elastoplastic pushover analysis method to evaluate the remaining seismic performance of the structure, but its accuracy in simulating the actual dynamic process of an earthquake is inferior to that of dynamic time history analysis methods.

[0113] This application proposes an assessment method based on the performance degradation of a complete structural system. Its core lies in generating training data using a two-stage seismic finite element simulation. The training data is generated through prior numerical simulation, and the selected data considers the structural damage state, avoiding reliance on a large amount of difficult-to-obtain field or experimental image data. The fiber beam model used in the numerical simulation has been verified through extensive experiments, demonstrating high computational reliability and ensuring the authenticity of the simulation data. In the application phase, the assessment process requires no finite element calculations, significantly reducing assessment time and achieving rapid end-to-end assessment output.

[0114] To address the bottlenecks in existing technologies, this invention proposes a deep learning-based method for assessing and deciding on the safety performance of post-earthquake frame structures, based on the need for rapid and accurate evaluation of their safety performance. This method integrates damage state data such as residual inter-story drift angles and plastic hinge ratios, and combines a long short-term memory neural network to capture the temporal characteristics of seismic waves. A hybrid MLP-LSTM architecture is designed to achieve composite feature learning that combines static damage features with dynamic seismic excitation. After verifying the model's generalization ability, this composite feature deep learning model can quickly predict the mechanical response of the structure under various seismic waves. Based on the damage level classification criteria, the probability distribution of damage levels after a re-earthquake is obtained, the structural safety performance is assessed, and maintenance decision indicators are calculated to make repair decisions. This enables rapid and accurate assessment and repair decisions for post-earthquake structural safety performance. In the assessment stage of post-earthquake frame structures, the mechanical response to seismic waves of different waveforms and intensities is considered to evaluate the current structural safety performance, thereby rationally allocating human and material resources for subsequent decisions such as repair or reconstruction.

[0115] In this embodiment of the invention, to meet the need for rapid and accurate seismic performance assessment of existing composite frame structures, a deep learning-based method for post-earthquake safety assessment of frame structures is proposed. During the assessment phase after an earthquake, the method considers the responses that can characterize the structural state generated by subsequent seismic waves of different waveforms and intensities. Based on the re-earthquake response, the current damage state of the structure is assessed, and repair decisions are made. This helps improve the efficiency and reliability of subsequent seismic performance assessment of damaged structures, promoting further development and application in the field of structural damage assessment.

[0116] Figure 8 This is a block diagram illustrating a post-earthquake structural assessment and repair apparatus based on two-stage time history analysis, according to an exemplary embodiment. The apparatus is used in a post-earthquake structural assessment and repair method based on two-stage time history analysis. (Refer to...) Figure 8 The device includes an evaluation module 310. Wherein:

[0117] The assessment module 310 is used to acquire the structural attribute parameter data and structural response data of the post-earthquake structure to be assessed, input the structural attribute parameter data and structural response data of the post-earthquake structure to be assessed into the constructed assessment model, and obtain the safety assessment results and repair decisions of the post-earthquake structure.

[0118] The process of constructing the evaluation model includes:

[0119] S1. Construct a multi-story structure based on structural property parameter data. Based on the collected seismic wave acceleration sequence data, perform multiple two-stage elastoplastic time history analyses on the multi-story structure using a fiber beam model. Simulate various working conditions of the multi-story structure under two earthquakes. Obtain the structural response data of the first stage and the corresponding structural response data of the second stage under various two-stage working conditions of the multi-story structure. Construct a training dataset based on the structural property parameter data, the structural response data of the first stage, the seismic wave acceleration sequence data of the second stage, and the structural response data of the second stage.

[0120] S2. A deep learning model is built based on a multilayer perceptron (MLP) and a long short-term memory (LSTM) neural network. The deep learning model is trained using the training dataset to obtain a trained deep learning model. The trained deep learning model has the ability to predict the structural response data of the second stage by inputting structural attribute parameter data, structural response data of the first stage, and seismic wave acceleration sequence data of the second stage.

[0121] S3. The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed to reduce the dimension of the data by using the evaluation index dimensionality reduction method. This yields one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic waves. Based on the dataset composed of the one-dimensional damage assessment indexes of the structure after experiencing various second-stage seismic waves, a probability density curve is constructed, and the 95th percentile of the probability density curve is obtained.

[0122] S4. Define the damage state, calculate the critical displacement angle between each level of damage state, and obtain the structural mechanical response state corresponding to the critical displacement angle between each level of damage state. The structural mechanical response state is characterized by the response data. The response data is weighted and summed to reduce the dimensionality of the evaluation index, and the one-dimensional damage evaluation index limit of the structural mechanical response state is obtained as the basis for damage level classification.

[0123] S5. Based on the 95th percentile of the probability density curve and the damage level classification criteria, the post-earthquake frame structure safety assessment results and repair decisions are obtained.

[0124] In this embodiment of the invention, to meet the need for rapid and accurate seismic performance assessment of existing composite frame structures, a deep learning-based method for post-earthquake safety assessment of frame structures is proposed. During the assessment phase after an earthquake, the method considers the responses that can characterize the structural state generated by subsequent seismic waves of different waveforms and intensities. Based on the re-earthquake response, the current damage state of the structure is assessed, and repair decisions are made. This helps improve the efficiency and reliability of subsequent seismic performance assessment of damaged structures, promoting further development and application in the field of structural damage assessment.

[0125] Figure 9 This is a structural schematic diagram of a post-earthquake structural assessment and repair device provided in an embodiment of the present invention, as shown below. Figure 9 As shown, post-earthquake structural assessment and repair equipment may include the above-mentioned... Figure 8 The illustrated post-earthquake structural assessment and repair device is based on two-stage time history analysis. Optionally, the post-earthquake structural assessment and repair device 410 may include a first processor 2001.

[0126] Optionally, the post-earthquake structural assessment and repair equipment 410 may also include a memory 2002 and a transceiver 2003.

[0127] The first processor 2001, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0128] The following is combined with Figure 9 A detailed description of each component of the post-earthquake structural assessment and repair equipment 410 is provided below:

[0129] The first processor 2001 is the control center of the post-earthquake structural assessment and repair equipment 410. It can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0130] Optionally, the first processor 2001 can perform various functions of the post-earthquake structural assessment and repair device 410 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0131] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, for example... Figure 9 CPU0 and CPU1 are shown in the diagram.

[0132] In a specific implementation, as one example, the post-earthquake structural assessment and repair device 410 may also include multiple processors, for example... Figure 9The first processor 2001 and the second processor 2004 are shown in the diagram. Each of these processors can be a single-core processor or a multi-core processor. Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (such as computer program instructions).

[0133] The memory 2002 is used to store the software program that executes the present invention, and is controlled by the first processor 2001 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0134] Optionally, the memory 2002 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 2002 may be integrated with the first processor 2001 or may exist independently, and may be connected via the interface circuit of the post-earthquake structural assessment and repair device 410. Figure 9 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0135] The transceiver 2003 is used to communicate with network devices or with terminal devices.

[0136] Alternatively, transceiver 2003 may include a receiver and a transmitter. Figure 9 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0137] Optionally, the transceiver 2003 can be integrated with the first processor 2001, or it can exist independently and be connected to the interface circuit of the post-earthquake structural assessment and repair equipment 410. Figure 9 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0138] It should be noted that, Figure 9 The structure of the post-earthquake structural assessment and repair device 410 shown in the diagram does not constitute a limitation on the router. The actual knowledge structure identification device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0139] Furthermore, the technical effect of the post-earthquake structural assessment and repair equipment 410 can be referred to the technical effect of the post-earthquake structural assessment and repair method based on two-stage time history analysis described in the above method embodiments, and will not be repeated here.

[0140] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0141] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0142] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0143] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0144] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0145] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0146] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0147] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0148] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0149] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0150] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0151] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for post-earthquake structural assessment and repair based on two-stage time history analysis, characterized in that, The method includes: Obtain structural attribute parameter data and structural response data of the post-earthquake structure to be evaluated, input the structural attribute parameter data and structural response data of the post-earthquake structure to be evaluated into the constructed evaluation model, and obtain the safety evaluation results and repair decisions of the post-earthquake structure. The process of constructing the evaluation model includes: S1. Construct a multi-story structure based on structural property parameter data. Based on the collected seismic wave acceleration sequence data, perform multiple two-stage elastoplastic time history analyses on the multi-story structure using a fiber beam model. Simulate various working conditions of the multi-story structure experiencing two earthquakes in succession to obtain the structural response data of the first stage and the corresponding structural response data of the second stage under various two-stage working conditions of the multi-story structure. Construct a training dataset based on the structural property parameter data, the structural response data of the first stage, the seismic wave acceleration sequence data of the second stage, and the structural response data of the second stage. S2. A deep learning model is built based on a multilayer perceptron (MLP) and a long short-term memory neural network (LSTM). The deep learning model is trained according to the training dataset to obtain a trained deep learning model. The trained deep learning model has the ability to predict the structural response data of the second stage by inputting structural attribute parameter data, structural response data of the first stage, and seismic wave acceleration sequence data of the second stage. S3. The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed to reduce the dimension by using the evaluation index dimensionality reduction method. This yields one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic waves. Based on the dataset of one-dimensional damage assessment indices of the structure after experiencing various second-stage seismic waves, a probability density curve is constructed, and the 95th quantile of the probability density curve is obtained. S4. Define the damage state, calculate the critical displacement angle between each level of damage state, and obtain the structural mechanical response state corresponding to the critical displacement angle between each level of damage state. The structural mechanical response state is characterized by the response data. The response data is weighted and summed to reduce the dimensionality of the evaluation index using the dimensionality reduction method to obtain the one-dimensional damage evaluation index limit of the structural mechanical response state as the basis for damage level classification. S5. Based on the 95th percentile of the probability density curve and the damage level classification criteria, the post-earthquake frame structure safety assessment results and repair decisions are obtained.

2. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, S1, based on the collected multiple seismic wave acceleration sequence data, uses a fiber beam model to perform multiple two-stage elastoplastic time history analyses on a multi-story structure, simulating various working conditions of the multi-story structure experiencing two earthquakes. This yields the first-stage structural response data and the corresponding second-stage structural response data for each of the two-stage working conditions, including: Based on the first-stage seismic wave acceleration sequence data, a first-stage elastoplastic time history analysis was performed on a multi-story structure using a fiber beam model to obtain the first-stage structural response data. The first-stage structural response data includes: residual inter-story drift angle data or plastic hinge rate data of each story in the first-stage damaged structure. Based on the second-stage seismic wave acceleration sequence data, a second-stage elastoplastic time history analysis was conducted on the multi-story structure after the first-stage earthquake using a fiber beam model to obtain the second-stage structural response data. The second-stage structural response data includes: the maximum inter-story drift angle data or the plastic hinge ratio data of each story of the earthquake-damaged structure in the second stage.

3. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, The second-stage structural response data, as predicted in S2 by inputting structural attribute parameter data, first-stage structural response data, and second-stage seismic wave acceleration sequence data, includes: The structural attribute parameter data and the structural response data of the first stage are input into the multilayer perceptron (MLP), and the output of the multilayer perceptron (MLP) is obtained after passing through three fully connected layers. The seismic wave acceleration sequence data of the second stage is input into the Long Short-Term Memory Neural Network (LSTM) to obtain the output of the LSTM. The outputs of the Multilayer Perceptron (MLP) and the Long Short-Term Memory (LSTM) Neural Network are passed through three fully connected layers to predict the structural response data for the second stage.

4. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, The dimensionality reduction method for the evaluation metrics in S3 includes: The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed according to a preset weight to reduce the dimension, thereby obtaining one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic wave. The calculation formula for the dimensionality reduction method of the evaluation index is shown in the following formula (1): （1） In the formula, This represents a one-dimensional damage assessment index of the structure after experiencing the second phase of the seismic wave. Represents the post-earthquake response matrix of the structure. Indicates matrix transpose. This represents the weight matrix.

5. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, The damage state defined in S4 includes: By performing lateral pushover analysis on a multi-layer structure, the load-displacement angle curve of the multi-layer structure under pushover action is obtained. The peak point and limit point of the load-displacement angle curve are determined, and the yield point of the load-displacement angle curve is determined by the farthest point method. Then, the damage state is defined. The damage state is used to describe the fully elastic, yield, peak and limit states of the multi-layer structure.

6. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, The calculation of the critical displacement angle between different levels of damage in S4 includes: The critical displacement angle between the first-level damage states is calculated as shown in equation (2) below: （2） The critical displacement angle between the second-level damage states is calculated as shown in equation (3): （3） The critical displacement angle between the third-level damage states is calculated as shown in equation (4): （4） In the formula, Indicates the critical displacement angle 1. This represents the displacement angle corresponding to the yield point. Indicates the critical displacement angle 2. This represents the displacement angle corresponding to the peak point. Indicates the critical displacement angle 3. This represents the displacement angle corresponding to the extreme point.

7. The method for post-earthquake structural assessment and repair based on two-stage time history analysis according to claim 1, characterized in that, The post-earthquake frame structure safety assessment results and repair decisions in S5, based on the 95th quantile of the probability density curve and the damage level classification criteria, include: When the 95th percentile of the probability density curve is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, the damage level is determined to be no damage, and the person can move in immediately. When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the first-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, the damage level is determined to be minor damage, and minor repairs are required before relocation. When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the second-level damage states, and is less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be moderate damage, and repairs are required before relocation. When the 95th percentile of the probability density curve is not less than the limit of the one-dimensional damage assessment index corresponding to the critical displacement angle between the third-level damage states, the damage level is determined to be severe damage, and the structure needs to be rebuilt.

8. A post-earthquake structural assessment and repair device based on two-stage time history analysis, wherein the device is used to implement the post-earthquake structural assessment and repair method based on two-stage time history analysis as described in any one of claims 1-7, characterized in that, The device includes: The assessment module is used to acquire the structural attribute parameter data and structural response data of the post-earthquake structure to be assessed, input the structural attribute parameter data and structural response data of the post-earthquake structure to be assessed into the constructed assessment model, and obtain the safety assessment results and repair decisions of the post-earthquake structure. The process of constructing the evaluation model includes: S1. Construct a multi-story structure based on structural property parameter data. Based on the collected seismic wave acceleration sequence data, perform multiple two-stage elastoplastic time history analyses on the multi-story structure using a fiber beam model. Simulate various working conditions of the multi-story structure experiencing two earthquakes in succession to obtain the structural response data of the first stage and the corresponding structural response data of the second stage under various two-stage working conditions of the multi-story structure. Construct a training dataset based on the structural property parameter data, the structural response data of the first stage, the seismic wave acceleration sequence data of the second stage, and the structural response data of the second stage. S2. A deep learning model is built based on a multilayer perceptron (MLP) and a long short-term memory neural network (LSTM). The deep learning model is trained according to the training dataset to obtain a trained deep learning model. The trained deep learning model has the ability to predict the structural response data of the second stage by inputting structural attribute parameter data, structural response data of the first stage, and seismic wave acceleration sequence data of the second stage. S3. The maximum inter-story drift angle data or plastic hinge rate data of each layer in the predicted second-stage structural response data are weighted and summed to reduce the dimension by using the evaluation index dimensionality reduction method. This yields one-dimensional damage assessment index data of the structure after experiencing the second-stage seismic waves. Based on the dataset of one-dimensional damage assessment indices of the structure after experiencing various second-stage seismic waves, a probability density curve is constructed, and the 95th quantile of the probability density curve is obtained. S4. Define the damage state, calculate the critical displacement angle between each level of damage state, and obtain the structural mechanical response state corresponding to the critical displacement angle between each level of damage state. The structural mechanical response state is characterized by the response data. The response data is weighted and summed to reduce the dimensionality of the evaluation index using the dimensionality reduction method to obtain the one-dimensional damage evaluation index limit of the structural mechanical response state as the basis for damage level classification. S5. Based on the 95th percentile of the probability density curve and the damage level classification criteria, the post-earthquake frame structure safety assessment results and repair decisions are obtained.

9. A post-earthquake structural assessment and repair device, characterized in that, The post-earthquake structural assessment and repair equipment includes: processor; A memory storing computer-readable instructions that, when executed by the processor, implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1 to 7.

Images