A seismic damage prediction method and system based on the Kriging model

Abstract

An embodiment of the present application provides a seismic damage prediction method and system based on the Kriging model. The method includes the following steps: obtaining model training data; using the obtained model training data to train the Kriging model to obtain a trained seismic capacity Kriging model and a seismic demand Kriging model; and obtaining a seismic damage prediction result based on the trained seismic capacity Kriging model and the seismic demand Kriging model. Based on the Kriging model, the present application uses the corresponding data set to train the Kriging model, and respectively obtains the seismic capacity and seismic demand of the structure through the trained Kriging model, so as to effectively predict the seismic damage that occurs to the structure under the specified seismic action.

Description

Technical Field

[0001] This application relates to the field of earthquake resistance technology in civil engineering, and specifically relates to a method and system for earthquake damage prediction based on the Kriging model. Background Art

[0002] After an earthquake occurs, information isolation often occurs between the disaster area and managers and rescuers due to severe damage to communication, transportation networks, etc. Accurately predicting the earthquake damage in the disaster area within a few hours after the earthquake is the key to ensuring the success of rescue operations.

[0003] As a surrogate model, the Kriging model can better handle the trade-off between the efficiency and accuracy of calculations, and has been widely used in computationally intensive problems such as structural optimization, risk prediction, sensitivity analysis, and structural reliability prediction.

[0004] In the prior art, there is a lack of an earthquake damage prediction method that can effectively balance calculation efficiency and calculation accuracy. Summary of the Invention

[0005] This application provides a method and system for earthquake damage prediction based on the Kriging model, which can solve the technical problem that the existing earthquake damage prediction method cannot effectively balance calculation efficiency and calculation accuracy.

[0006] In the first aspect, this application provides a method for earthquake damage prediction based on the Kriging model, including the following steps:

[0007] Obtain model training data;

[0008] Use the obtained model training data to train the Kriging model to obtain the trained earthquake resistance ability Kriging model and earthquake resistance demand Kriging model;

[0009] Based on the trained earthquake resistance ability Kriging model and earthquake resistance demand Kriging model, obtain the earthquake damage prediction result.

[0010] Combined with the first aspect, in one implementation, the step of obtaining model training data specifically includes the following steps:

[0011] Obtain ground motion parameters;

[0012] Obtain the time-history analysis structural response data of ground motion records.

[0013] Combined with the first aspect, in one implementation, the obtaining of the time-history analysis structural response data of ground motion records includes the following steps:

[0014] Build a finite element model of the structure;

[0015] Perform elastoplastic time history analysis on the obtained ground motion records and the finite element model of the constructed structure to obtain the time history analysis structural response data.

[0016] Combined with the first aspect, in one implementation, the steps of training the Kriging model with the obtained model training data to obtain the trained seismic capacity Kriging model and seismic demand Kriging model specifically include the following steps:

[0017] Use the data of base shear - maximum interstory drift ratio obtained from the time history analysis structural response data to train the Kriging model to obtain the trained seismic capacity Kriging model;

[0018] Use the time history analysis structural response data and ground motion parameters to train the Kriging model to obtain the trained seismic demand Kriging model.

[0019] Combined with the first aspect, in one implementation, the steps of using the data of base shear - maximum interstory drift ratio obtained from the time history analysis structural response data to train the Kriging model to obtain the trained seismic capacity Kriging model specifically include the following steps:

[0020] Define the maximum interstory drift ratios for different stiffness intervals;

[0021] Use the time history analysis structural response data to train the Kriging model to obtain the seismic capacity curve, and according to the definition of the earthquake damage level, obtain the maximum interstory drift ratio limits corresponding to each earthquake damage level.

[0022] Combined with the first aspect, in one implementation, the steps of using the time history analysis structural response data and ground motion parameters to train the Kriging model to obtain the trained seismic demand Kriging model specifically include the following steps:

[0023] Divide the obtained time history analysis data into a training set and a test set, use the training set data to train the Kriging model, and input the ground motion parameters into the trained Kriging model to obtain the seismic demand Kriging model corresponding to the frame structure.

[0024] Combined with the first aspect, in one implementation, after the steps of training the Kriging model with the obtained model training data to obtain the trained seismic capacity Kriging model and seismic demand Kriging model, the following steps are further included:

[0025] Verify the earthquake damage prediction of the trained earthquake resistance capacity Kriging model and the earthquake resistance demand Kriging model, and obtain the earthquake damage prediction verification results.

[0026] In a second aspect, the present application provides an earthquake damage prediction system based on a Kriging model, including:

[0027] A training data acquisition module for acquiring model training data;

[0028] A model training module, communicatively connected to the training data acquisition module, for training a Kriging model using the acquired model training data to obtain a trained earthquake resistance capacity Kriging model and an earthquake resistance demand Kriging model;

[0029] An earthquake damage prediction module, communicatively connected to the model training module, for obtaining earthquake damage prediction results based on the trained earthquake resistance capacity Kriging model and the earthquake resistance demand Kriging model.

[0030] In combination with the second aspect, in an implementation manner, the training data acquisition module includes:

[0031] An earthquake motion parameter acquisition unit for acquiring earthquake motion parameters;

[0032] And a time history analysis data acquisition unit for acquiring the time history analysis structural response data of earthquake motion records.

[0033] The beneficial effects brought by the technical solutions provided in the embodiments of the present application at least include:

[0034] An earthquake damage prediction method based on a Kriging model provided by the present application is based on the Kriging model, trains the Kriging model using corresponding data sets, and respectively obtains the earthquake resistance capacity and earthquake resistance demand of the structure through the trained Kriging model, so as to effectively predict the earthquake damage that occurs to the structure under a specified earthquake action. Description of the Drawings

[0035] Figure 1 It is a schematic flowchart of the earthquake damage prediction method based on the Kriging model provided in the embodiment of the present application;

[0036] Figure 2 It is the earthquake damage prediction flowchart of the earthquake damage prediction method based on the Kriging model provided in the embodiment of the present application;

[0037] Figure 3 It is the statistical histogram of earthquake motion parameters of the earthquake damage prediction method based on the Kriging model provided in the embodiment of the present application;

[0038] Figure 4 The floor plan of the framework structure of the earthquake damage prediction method based on the Kriging model provided by the embodiment of the present application;

[0039] Figure 5 The refined modeling diagram of the framework structure of the earthquake damage prediction method based on the Kriging model provided by the embodiment of the present application;

[0040] Figure 6 The statistical histogram of earthquake response of the earthquake damage prediction method based on the Kriging model provided by the embodiment of the present application;

[0041] Figure 7 The typical earthquake damage scenario diagram of the framework structure of the earthquake damage prediction method based on the Kriging model provided by the embodiment of the present application. Detailed implementation manners

[0042] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0043] The terms "including" and "having" and any variations thereof in the description of the specification, claims and drawings of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices. The descriptions with terms such as "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit that "first", "second" and "third" are of different types.

[0044] In the description of the embodiments of the present application, terms such as "exemplary", "for example" or "for instance" are used to indicate examples, illustrations or explanations. Any embodiment or design solution described as "exemplary", "for example" or "for instance" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly, the use of words such as "exemplary", "for example" or "for instance" is intended to present related concepts in a specific manner.

[0045] In the description of the embodiments of the present application, unless otherwise specified, " / " means "or". For example, A / B may mean A or B. The "and / or" in the text is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B may mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present application, "a plurality of" means two or more than two.

[0046] In some processes described in the embodiments of the present application, a plurality of operations or steps appear in a specific order. However, it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or may be executed in parallel. The serial numbers of the operations are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in order or in parallel, and these operations or steps may be combined.

[0047] In a first aspect, please refer to Figure 1 , the present application provides a seismic damage prediction method based on the Kriging model, including the following steps:

[0048] Step S1, obtain model training data;

[0049] Step S2, use the obtained model training data to train the Kriging model to obtain the trained Kriging model of seismic capacity and the Kriging model of seismic demand;

[0050] Step S3, based on the trained Kriging model of seismic capacity and the Kriging model of seismic demand, obtain the seismic damage prediction result.

[0051] The seismic damage prediction method based on the Kriging model provided by the present application is based on the Kriging model, uses the corresponding data set to train the Kriging model, and obtains the seismic capacity and seismic demand of the structure through the trained Kriging model, so as to effectively predict the seismic damage that occurs to the structure under the specified seismic action.

[0052] In one embodiment, as Figure 2 shown, the step S1, the step of obtaining model training data, specifically includes the following steps:

[0053] Step S11, obtain ground motion parameters;

[0054] Step S12, obtain the time history analysis structural response data of the ground motion record.

[0055] In one embodiment, the step S11 of obtaining ground motion parameters specifically includes the following steps:

[0056] Obtain ground motion records, and extract ground motion parameters from the obtained ground motion records. Specifically, three types of ground motion parameters are selected: pulse-related parameters, acceleration-related parameters, and velocity-related parameters. A total of 17 commonly used ground motion parameters are used as the initial input parameters of the Kriging model for structural response.

[0057] The multivariate ground motion parameters and calculation methods are shown in Table 1, and the parameter statistical histograms of the ground motion records used are as Figure 3 shown.

[0058] Figure 3 as follows:

[0059] Figure 3 (a) is the frequency distribution histogram of the pulse period;

[0060] Figure 3 (b) is the frequency distribution histogram of the pulse factor;

[0061] Figure 3 (c) is the frequency distribution histogram of the peak ground motion velocity;

[0062] Figure 3 (d) is the frequency distribution histogram of the peak ground motion acceleration;

[0063] Figure 3 (e) is the frequency distribution histogram of the spectral acceleration of the first period;

[0064] Figure 3 (f) is the frequency distribution histogram of the peak of the acceleration response spectrum;

[0065] Figure 3 (g) is the frequency distribution histogram of the effective peak acceleration;

[0066] Figure 3 (h) is the frequency distribution histogram of the Riddell index Ia;

[0067] Figure 3 (i) is the frequency distribution histogram of the Housner intensity;

[0068] Figure 3 (j) is the frequency distribution histogram of the Arias intensity;

[0069] Figure 3 (k) is the frequency distribution histogram of the Park-Ang index;

[0070] Figure 3 (l) is the frequency distribution histogram of the maximum difference between positive and negative peak velocities;

[0071] Figure 3 (m) is the frequency distribution histogram of the peak value of the velocity response spectrum;

[0072] Figure 3 (n) is the frequency distribution histogram of the Housner intensity;

[0073] Figure 3 (o) is the frequency distribution histogram of the modified Arias intensity;

[0074] Figure 3 (p) is the frequency distribution histogram of the Riddell index Iv;

[0075] Figure 3 (q) is the frequency distribution histogram of the Faifar index IF.

[0076] Table 1 Multivariate parameters of ground motion and calculation methods

[0077]

[0078] Note: In the table, PGV ratio is the ratio of the PGV of the remaining waveform signal after extracting the pulse signal to the PGV of the original record; E ratio is the ratio of the energy of the remaining waveform signal to the energy of the original record. When this ratio is greater than 0.85, it is pulse type, and when it is less than 0.15, it is non - pulse type; and are the acceleration and velocity time histories of the ground motion record respectively; S a and S v are the response spectrum acceleration and velocity of the ground motion record respectively; T is the natural vibration period of the structure; t5 and t 95 are the corresponding times when the calculated Arias intensities account for 5% and 95% of the calculated Arias intensity at the end of the entire earthquake respectively; T f is the total duration of the ground motion; v0 is the number of times the ground motion acceleration curve passes through zero per unit time.

[0079] In one embodiment, the step S12 of obtaining the time - history analysis structural response data of the ground motion record includes the following steps:

[0080] Step S121, constructing a finite - element model of the structure;

[0081] In view of the fact that a certain museum on the Moxi Terrace in Luding County was severely damaged in the magnitude 6.8 earthquake that occurred on September 5, 2022, this application selects the frame structure with the most severe earthquake damage in this museum for analysis and verifies the earthquake damage prediction method for this frame structure. The floor plan of the frame structure is as Figure 4 shown. Figure 4(a) is the floor plan of the frame structure of the first floor, Figure 4 (b) is the floor plan of the frame structure of the second floor.

[0082] Step S122: Perform elastoplastic time history analysis on the obtained ground motion records and the finite element model of the constructed structure to obtain the structural response data of the time history analysis:

[0083] This application uses the general analysis software ANSYS to carry out refined modeling of the frame structure. As Figure 5 shown, it is a frame structure of the second floor, and the heights of the first floor and the second floor are 4.8m and 4.2m respectively;

[0084] According to the on-site rebound test and test results, the concrete strength of the frame columns, frame beams and floor slabs is C30, the steel bars are HRB400, and the strength values of various materials are all standard values. The concrete and steel bars respectively adopt multi-segment and double-segment ideal elastoplastic constitutive relations, and both adopt isotropic hardening models, without considering the Bauschinger effect. The frame beams and "imitated bracket" are modeled by 3D linear finite strain beam element BEAM188; the floor slabs are modeled by 3D finite strain shell element SHELL181; in order to accurately account for the influence of stirrups on the plastic development of frame columns, within the range of 1.5m at the bottom of the first-floor frame columns, 3D solid elements with reinforcement SOLID65 are used, and the remaining frame columns are simulated by 3D quadratic finite strain beam element BEAM189; the longitudinal steel bars of the components are all dispersed in the entire cross-section. The entity element part at the lower end of the column is all divided by hexahedron mapping and the mesh is encrypted. The elastoplastic time history analysis considers the geometric nonlinearity of the frame structure, and the number of meshes of the frame structure is 26464.

[0085] The structural responses obtained from the elastoplastic time history analysis of the finite element model are respectively: the inter-story drift angle of the first floor, the inter-story drift angle of the second floor and the base shear force. The statistical histograms of the structural responses are shown in Figure 6 .

[0086] Figure 6 as follows:

[0087] Figure 6 (a) is the frequency distribution histogram of the maximum inter-story drift angle of the first floor;

[0088] Figure 6 (b) is the frequency distribution histogram of the maximum inter-story drift angle of the second floor;

[0089] Figure 6 (c) is the frequency distribution histogram of the total base shear force.

[0090] In an embodiment, the step S2: using the obtained model training data to train the Kriging model to obtain the trained Kriging model of seismic capacity and the Kriging model of seismic demand specifically includes the following steps:

[0091] Step S21: Use the data of base shear - maximum inter - story drift ratio obtained from the structural response data of time - history analysis to train the Kriging model, and obtain the trained Kriging model of seismic capacity.

[0092] Step S22: Use the structural response data of time - history analysis and ground motion parameters to train the Kriging model, and obtain the trained Kriging model of seismic demand.

[0093] In one embodiment, the step S21: Use the data of base shear - maximum inter - story drift ratio obtained from the structural response data of time - history analysis to train the Kriging model, and obtain the trained Kriging model of seismic capacity specifically includes the following steps:

[0094] Step S211: Define the maximum inter - story drift ratios in different stiffness intervals, specifically:

[0095] A. Define the maximum inter - story drift ratio corresponding to the situation where the overall stiffness of the structure drops to 90% of the initial stiffness as the yield inter - story drift ratio of the structure.

[0096] B. Define the maximum inter - story drift ratio corresponding to the situation where the overall stiffness of the structure drops to 70% of the initial stiffness as the elastic - plastic inter - story drift ratio of the structure.

[0097] C. Define the maximum inter - story drift ratio corresponding to the situation where the overall stiffness of the structure drops to 20% of the initial stiffness as the ultimate inter - story drift ratio of the structure.

[0098] Define the maximum limit value of the inter - story drift ratio that the structure can reach as the critical displacement angle of the structure. When the maximum inter - story drift ratio exceeds this limit value, the structure will collapse.

[0099] According to the definitions of the stiffness intervals of the three maximum inter - story drift ratios A, B, and C above, the capacity curve of the structure can be divided into five earthquake damage grade sections:

[0100] (1) Initial state - yield state, corresponding earthquake damage grade is "undamaged";

[0101] (2) Yield state - elastic - plastic state, corresponding earthquake damage grade is "slightly damaged";

[0102] (3) Elastic - plastic state - ultimate state, corresponding earthquake damage grade is "moderately damaged";

[0103] (4) Ultimate state - critical state, corresponding earthquake damage grade is "severely damaged";

[0104] (5) After exceeding the critical state, local or overall collapse of the structure occurs.

[0105] Step S212: Use the time history analysis structural response data to train the Kriging model to obtain the seismic capacity curve. According to the definition of the earthquake damage level, obtain the maximum inter-story drift angle limits corresponding to each earthquake damage level, as shown in Table 2:

[0106] Table 2 Maximum inter-story drift angle limits corresponding to the earthquake damage level of Frame W2

[0107]

[0108] Note: μ and σ are the mean and standard deviation of the capacity curve, respectively.

[0109] In one embodiment, the steps of using the time history analysis structural response data and ground motion parameters to train the Kriging model to obtain the trained seismic demand Kriging model specifically include the following steps:

[0110] Divide the obtained time history analysis data into a training set and a test set. More specifically, divide the results of the time history analysis into a training set and a test set according to a ratio of 80%:20%. Use the training set data to train the Kriging model, use the test set to test the trained Kriging model, and input the ground motion parameters into the trained Kriging model to obtain the seismic demand Kriging model corresponding to the frame structure. The seismic response simulation results of the Kriging model are shown in Table 3:

[0111] Table 3 Simulation results of the seismic demand Kriging model

[0112]

[0113] In one embodiment, determine the training method of the Kriging model;

[0114] The Kriging model is a parameterized "input-output" system and can replace the finite element model in earthquake damage prediction to improve the efficiency and accuracy of earthquake damage prediction. For the system y(X) = g(X), the mathematical expression of the Kriging model can be written as:

[0115] g(X) = f T (X)β + z(X) (1)

[0116] In the formula, y(X) is the output parameter; g(X) is the Kriging model to be trained;

[0117] f T (X) = [f1(X), f2(X), …, f p (X)] T is the basis function of the input variable X, and P is the number of basis functions;

[0118] β = [β1, β2, …, β p T are undetermined parameters estimated based on the training set; z(X) is a stochastic process of the local deviation of the Kriging model with an expectation of 0 and a variance of

[0119] The covariance of z(X) in Equation (1) can be written as:

[0120]

[0121] where R(Θ, x v , x w ) is the correlation function of the sample points x v and x w , v, w = 1, 2, …, m, where m is the number of samples in the training set; Θ is the hyperparameter of the model, Θ = [θ1, …, θ i , …, θ l T .

[0122] In this application, the Matern type is selected as the form of the correlation function, then R(Θ, x v , x w ) can be expressed as

[0123]

[0124] where ||x v - x w || is the Euclidean distance; K v (·) is the modified Bessel function; Γ(·) is the Gamma function; the parameter ν controls the smoothness of the correlation function, and in this application, it takes the value of 1.5; the parameter Θ is the characteristic length scale parameter, and the smaller the characteristic length scale, the higher the correlation between samples. To improve the robustness of the Kriging model, Θ can be optimized as a model parameter.

[0125] The estimated value of the undetermined parameter β in Equation (1) can be calculated according to the following formula:

[0126]

[0127] where R is the correlation function matrix corresponding to m training samples; F is an m×p matrix composed of the values of P basis functions at all m training sample points; g is a column vector composed of the output values of all training set samples.

[0128] The variance of z(X) of the estimated value can be estimated according to the least squares theory and the following formula:​​

[0129]

[0130] In this application, the hyperparameter Θ = [θ1, …, θ i , …, θ p is calculated through maximum likelihood estimation T , and can be expressed by the following formula:

[0131]

[0132] The training process of the Kriging model is the process of solving the hyperparameter Θ and the parameter L. Since there is no analytical solution for Θ and L, gradient algorithms such as the quasi - Newton method or gradient - free algorithms such as genetic algorithms need to be used. The specific training method of the Kriging model should be selected in combination with the data. After the Kriging model is trained, for the unknown sample x′, z(x′) can be calculated according to the following formula:

[0133]

[0134] In the formula, r T (x′) is the correlation between a single sample and the training samples,

[0135] r T (x′) = [R(x′, x1), R(x′, x2), …, R(x′, x m )] T .

[0136] According to Equation (1) and Equation (7), the best unbiased estimate of the output result of the Kriging model for the unknown sample x′ can be written as

[0137]

[0138] For the trained Kriging model, this application uses the coefficient of determination R 2 as the evaluation index of the accuracy of its simulation results. The expression of the coefficient of determination R 2 is:

[0139]

[0140] In the formula, y(x k ) is the actual output result corresponding to the sample x k , and is the average value of the actual output results.

[0141] In one embodiment, after the step of training the Kriging model using the obtained model training data to obtain the trained Kriging model for seismic capacity and the Kriging model for seismic demand, the following steps are further included:

[0142] Perform seismic damage prediction verification on the trained Kriging model for seismic capacity and the Kriging model for seismic demand to obtain the seismic damage prediction verification results.

[0143] In one embodiment, after the step of training the Kriging model using the obtained model training data to obtain the trained Kriging model for seismic capacity and the Kriging model for seismic demand, the following steps are further included:

[0144] Perform seismic damage prediction verification on the trained Kriging model for seismic capacity and the Kriging model for seismic demand to obtain the seismic damage prediction verification results.

[0145] In the Luding 6.8-magnitude earthquake that occurred on September 5, 2022, the frame in the present invention collapsed: a large number of plastic hinges appeared at the upper and lower ends of the first-floor frame columns, the second-floor floor slab underwent overall sliding, the roof floor slab cracked severely, and the overall structure collapsed. Typical seismic damages are shown in Figure 7 .

[0146] Figure 7 as follows:

[0147] Figure 7 (a)- Figure 7 (c) are typical seismic damage diagrams of the beam-column joints on the first floor;

[0148] Figure 7 (d)- Figure 7 (f) are typical seismic damage diagrams of the column bases on the first floor;

[0149] Figure 7 (d) is a seismic damage diagram of shear failure;

[0150] Figure 7 (e) is a seismic damage diagram of compression failure;

[0151] Figure 7 (f) is a seismic damage diagram of flexural failure;

[0152] Figure 7 (g)- Figure 7 (h) are seismic damage diagrams of inclination;

[0153] Figure 7 (g) is a seismic damage diagram of the inclination of the first-floor floor slab;

[0154] Figure 7 (h) is a seismic damage diagram of the inclination of the second-floor roof.

[0155] In the present invention, the accuracy of the earthquake damage prediction method is verified by using the time history analysis results of the finite element model and the actual earthquake damage investigation results respectively. Considering the uncertainty of the structural seismic response, the value of the maximum inter-story drift angle μ + 2σ is used as the final calculation result. The results of predicting the earthquake damage of the frame structure in the Luding earthquake by using the earthquake damage prediction method based on the Kriging model are shown in Table 4.

[0156] Table 4 Time history analysis results of the finite element model and earthquake damage prediction results of the earthquake damage prediction method based on the Kriging model provided in this application

[0157] Simulation method Maximum inter-story drift ratio Base shear / kN Performance level Elasto-plastic time history analysis 1 / 101 7825 Collapse Kriging model 1 / 108 11780 Collapse

[0158] In a second aspect, this application provides an earthquake damage prediction system based on the Kriging model, including a training data acquisition module, a model training module, and an earthquake damage prediction module. The training data acquisition module is used to acquire model training data; the model training module is communicatively connected to the training data acquisition module and is used to train the Kriging model by using the acquired model training data to obtain the trained Kriging model of seismic capacity and the Kriging model of seismic demand; the earthquake damage prediction module is communicatively connected to the model training module and is used to obtain earthquake damage prediction results based on the trained Kriging model of seismic capacity and the Kriging model of seismic demand.

[0159] In one embodiment, the training data acquisition module includes:

[0160] An earthquake motion parameter acquisition unit, which is used to acquire earthquake motion parameters;

[0161] A time history analysis data acquisition unit, which is used to acquire the time history analysis structural response data of earthquake motion records.

[0162] In a third aspect, an embodiment of this application also provides a readable storage medium.

[0163] The readable storage medium of this application stores an earthquake damage prediction program based on the Kriging model. When the earthquake damage prediction program based on the Kriging model is executed by a processor, the steps of the earthquake damage prediction method based on the Kriging model as described above are implemented.

[0164] Among them, the method implemented when the earthquake damage prediction program based on the Kriging model is executed can refer to various embodiments of the earthquake damage prediction method based on the Kriging model of this application, which will not be elaborated here.

[0165] It should be noted that the serial numbers of the above embodiments of this application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0166] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above and includes several instructions for causing a terminal device to execute the methods described in various embodiments of the present application.

[0167] The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A seismic damage prediction method based on the Kriging model, characterized in that It includes the following steps: Obtain model training data; Use the obtained model training data to train the Kriging model to obtain the trained seismic capacity Kriging model and seismic demand Kriging model; Based on the trained seismic capacity Kriging model and seismic demand Kriging model, obtain the earthquake damage prediction result; The step of obtaining model training data specifically includes the following steps: Obtain ground motion parameters, where the ground motion parameters include 3 pulse-related parameters, 8 acceleration-related parameters, and 6 velocity-related parameters; Obtain the time history analysis structural response data of the ground motion record; The step of obtaining the time history analysis structural response data of the ground motion record includes the following steps: Construct a finite element model of the structure; Perform elastoplastic time history analysis on the obtained ground motion record and the constructed finite element model of the structure to obtain the time history analysis structural response data; The step of using the obtained model training data to train the Kriging model to obtain the trained seismic capacity Kriging model and seismic demand Kriging model specifically includes the following steps: Use the data of base shear-maximum inter-story drift ratio obtained from the time history analysis structural response data to train the Kriging model to obtain the trained seismic capacity Kriging model; Use the time history analysis structural response data and ground motion parameters to train the Kriging model to obtain the trained seismic demand Kriging model; The step of using the data of base shear-maximum inter-story drift ratio obtained from the time history analysis structural response data to train the Kriging model to obtain the trained seismic capacity Kriging model specifically includes the following steps: Define the maximum inter-story drift ratios of different stiffness intervals: A. Define the maximum inter-story drift ratio corresponding to the overall stiffness of the structure dropping to 90% of the initial stiffness as the yield inter-story drift ratio of the structure; B. Define the maximum inter-story drift ratio corresponding to the overall stiffness of the structure dropping to 70% of the initial stiffness as the elastoplastic inter-story drift ratio of the structure; C. Define the maximum inter-story drift ratio corresponding to the overall stiffness of the structure dropping to 20% of the initial stiffness as the ultimate inter-story drift ratio of the structure; Define the maximum allowable inter-story drift ratio that the structure can reach as the critical displacement angle of the structure. When the maximum inter-story drift ratio exceeds this limit, the structure will collapse; According to the stiffness interval definitions of the maximum inter-story drift ratios in A, B, and C, divide the capacity curve of the structure into five earthquake damage grade sections: (1) Starting state - yield state, corresponding earthquake damage grade is "no damage"; (2) Yield state - elastoplastic state, corresponding earthquake damage grade is "slight damage"; (3) Elastoplastic state - ultimate state, corresponding earthquake damage grade is "moderate damage"; (4) Ultimate state - critical state, corresponding earthquake damage grade is "severe damage"; (5) After exceeding the critical state, local or overall collapse of the structure occurs; Using the structural response data from time - history analysis to train the Kriging model to obtain the seismic capacity curve, and according to the definition of the seismic damage level, obtaining the maximum inter - story drift angle limit values corresponding to each seismic damage level.

2. The earthquake damage prediction method based on the Kriging model according to claim 1, wherein, The steps of using the structural response data and ground motion parameters from time - history analysis to train the Kriging model to obtain the trained Kriging model of seismic demand specifically include the following steps: Divide the obtained time - history analysis data into a training set and a test set. Use the training set data to train the Kriging model, and input the ground motion parameters into the trained Kriging model to obtain the Kriging model of seismic demand corresponding to the frame structure.

3. The earthquake damage prediction method based on the Kriging model according to claim 1, characterized in that, After the steps of using the obtained model training data to train the Kriging model to obtain the trained Kriging model of seismic capacity and the trained Kriging model of seismic demand, the following steps are also included: Conduct seismic damage prediction verification on the trained Kriging model of seismic capacity and the trained Kriging model of seismic demand to obtain the seismic damage prediction verification results.

4. A seismic damage prediction system based on the Kriging model, characterized in that, Including: A training data acquisition module for acquiring model training data; A model training module, communicatively connected to the training data acquisition module, for using the obtained model training data to train the Kriging model to obtain the trained Kriging model of seismic capacity and the trained Kriging model of seismic demand; A seismic damage prediction module, communicatively connected to the model training module, for obtaining the seismic damage prediction results based on the trained Kriging model of seismic capacity and the trained Kriging model of seismic demand; The specific process of obtaining the model training data includes: Obtaining ground motion parameters; Obtaining the structural response data from time - history analysis of ground motion records; The process of obtaining the structural response data from time - history analysis of ground motion records includes: Constructing a finite - element model of the structure; Performing elastoplastic time - history analysis on the obtained ground motion records and the constructed finite - element model of the structure to obtain the structural response data from time - history analysis; The process of using the obtained model training data to train the Kriging model to obtain the trained Kriging model of seismic capacity and the trained Kriging model of seismic demand includes: Using the data of base shear - maximum inter - story drift angle obtained from the structural response data of time - history analysis to train the Kriging model to obtain the trained Kriging model of seismic capacity; Using the structural response data and ground motion parameters from time - history analysis to train the Kriging model to obtain the trained Kriging model of seismic demand; The steps of using the data of base shear - maximum inter - story drift angle obtained from the structural response data of time - history analysis to train the Kriging model to obtain the trained Kriging model of seismic capacity specifically include the following steps: Defining the maximum inter - story drift angle in different stiffness intervals: A. Defining the maximum inter - story drift angle corresponding to the overall structural stiffness dropping to 90% of the initial stiffness as the yield inter - story drift angle of the structure; B. Define the maximum inter-story drift angle corresponding to the overall structural stiffness reduced to 70% of the initial stiffness as the elastic-plastic inter-story drift angle of the structure; C. Define the maximum inter-story drift angle corresponding to the overall structural stiffness reduced to 20% of the initial stiffness as the ultimate inter-story drift angle of the structure; Define the maximum allowable inter-story drift angle limit that the structure can reach as the critical displacement angle of the structure. When the maximum inter-story drift angle exceeds this limit, the structure will collapse; According to the stiffness interval definitions of the three maximum inter-story drift angles A, B, and C, divide the capacity curve of the structure into five seismic damage level sections: (1) Initial state - Yield state, corresponding seismic damage level is "no damage"; (2) Yield state - Elastic-plastic state, corresponding seismic damage level is "minor damage"; (3) Elastic-plastic state - Ultimate state, corresponding seismic damage level is "moderate damage"; (4) Ultimate state - Critical state, corresponding seismic damage level is "severe damage"; (5) After exceeding the critical state, local or overall collapse of the structure occurs; Use time history analysis of structural response data to train the Kriging model to obtain the seismic capacity curve, and according to the definition of the seismic damage level, obtain the maximum inter-story drift angle limit corresponding to each seismic damage level.

Images