CN119126200A

CN119126200A - A method and device for selecting earthquake records in elastic time history analysis

Info

Publication number: CN119126200A
Application number: CN202411213727.8A
Authority: CN
Inventors: 任靖哲; 邹启令; 李治
Original assignee: CITIC General Institute of Architectural Design and Research Co Ltd
Current assignee: CITIC General Institute of Architectural Design and Research Co Ltd
Priority date: 2024-08-31
Filing date: 2024-08-31
Publication date: 2024-12-13
Anticipated expiration: 2044-08-31
Also published as: CN119126200B

Abstract

The invention relates to a method and a device for regulating and selecting a seismic record in elastic time course analysis, wherein the regulating and selecting method comprises the steps of adopting effective ground peak acceleration EPA to modulate amplitude of an initial seismic record adopted in the elastic time course analysis, and screening the initial seismic record according to a preset first regulating and selecting parameter to obtain a first seismic record; adopting causal inference to screen out seismic parameters which directly cause shearing force of the structural substrate as second tuning parameters; and screening the first seismic records according to the second screening parameters to obtain final seismic records for time course analysis. By the method, the discreteness of the shearing force of the final structure substrate can be well controlled, so that the correction of the vibration mode decomposition reaction spectrum result by using the elastic time-course analysis result is realized.

Description

Method and device for selecting earthquake records in elastic time course analysis

Technical Field

The invention relates to the technical field of structural earthquake resistance, in particular to a method and a device for selecting earthquake records in elastic time course analysis based on causal inference.

Background

According to relevant regulations, time-course analysis is needed for carrying out supplementary calculation under multiple earthquakes for particularly irregular buildings, class A buildings and buildings with high overrun. In the prior art, the effective ground peak acceleration EPA is adopted to tune and select the earthquake motion record, and although the earthquake motion record after tuning and selecting can be controlled to be approximately at the same intensity level with the designed reaction spectrum, even if part of the earthquake motion record is required to be manually recorded, the discreteness of the shearing force of the final structure substrate can not be well controlled, so that the result calculated by the elasticity time course analysis is often far beyond the result calculated by the vibration type decomposition reaction spectrum method, and the elasticity time course analysis result can not be used for correcting the reaction spectrum result. Therefore, the seismic record tuning method capable of controlling the shear dispersion of the substrate is required to be adopted in the elastic time course analysis and supplement calculation.

The approaches described in this section are not necessarily approaches that have been previously conceived or pursued. Unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section. Similarly, the problems mentioned in this section should not be considered as having been recognized in any prior art unless otherwise indicated.

Disclosure of Invention

Aiming at the technical problems in the prior art, the invention provides a seismic record tuning method in elastic time course analysis based on causal inference, which can be used for tuning seismic records in elastic time course checking.

The technical scheme for solving the technical problems is as follows:

in a first aspect, the present invention provides a method for tuning seismic records in elastic time-course analysis, including:

Amplitude modulation is carried out on the initial earthquake motion record adopted in the elastic time course analysis by adopting the effective ground peak acceleration EPA, and the initial earthquake motion record is screened according to a preset first tuning parameter to obtain a first earthquake motion record;

Adopting causal inference to screen out seismic parameters which directly cause shearing force of the structural substrate as second tuning parameters;

and screening the first seismic records according to the second screening parameters to obtain final seismic records for time course analysis.

Further, the method includes screening the initial seismic records according to a preset first tuning parameter to obtain first seismic records, wherein the initial seismic records with the difference between the average response spectrum of the screened amplitude-modulated initial seismic records and the designed response spectrum at the main structure periodic point of not more than 20% are used as the first seismic records.

Further, the method adopts causal inference to screen out the earthquake motion parameters which directly cause the shearing force of the structural substrate, and the earthquake motion parameters are used as second tuning parameters, and comprise:

extracting the seismic parameters contained in the first seismic record as candidate parameters for causal inference;

carrying out causal inference between the alternative parameters and the structural substrate shear force by adopting a potential result model;

Quantitatively analyzing the causal relation between the alternative parameters and the substrate shearing force under a potential result model frame by using a double-machine learning method to obtain causal effect indexes of each alternative parameter on the substrate shearing force;

And selecting an alternative parameter with the largest causal effect index as a second tuning parameter.

Further, the causal inference between the candidate parameter and the structural substrate shear force using the potential result model includes:

Assuming that the alternative parameters include two types of cause variable T and confusion variable X, the causal relationship between the structural substrate shear and the alternative parameters is expressed as follows:

Y=θT+g(X)+;

T=f(X)+;

Wherein Y is the substrate shear force generated after the intervention of the alternative parameter, g (X) and f (X) are unknown functions related to X, E and eta are unknown random errors with the mean value of 0, theta is the causal effect of the alternative parameter on the substrate shear force, theta = Y di (T = 1) ] -do (T = 0), T = 1 and T = 0 respectively represent the intervention/non-intervention, Y do (T = 1) represents the potential substrate shear force result generated by the earthquake record under the intervention of the reason variable T, and Y do (T = 0) represents the potential substrate shear force result generated by the earthquake record under the non-intervention.

Further, the method for quantitatively analyzing the causal relation between the candidate parameters and the substrate shearing force under the potential result model frame by using the double machine learning method to obtain causal effect indexes of each candidate parameter on the substrate shearing force comprises the following steps:

the substrate shearing force Y and the reason variable T generated after the intervention of the alternative parameters are respectively fitted by utilizing a XGBoost algorithm, and the optimal super parameters of the XGBoost algorithm are searched by adopting a Bayesian optimization algorithm to obtain estimated values of the substrate shearing force Y and the reason variable T And

Calculating fitting residual errors:

performing linear regression with the intercept of 0 on the fitting residual error, and calculating the estimated value of the causal effect theta generated by the alternative parameter on the substrate shearing force

ε_Y＝ε_T+;

Where n is the total amount of data in the designated training data set.

Further, the method for quantitatively analyzing the causal relationship between the candidate parameters and the substrate shearing force under the potential result model frame by using the double machine learning method to obtain causal effect indexes of each candidate parameter on the substrate shearing force further comprises:

The first seismic record D was randomly divided into b sub-datasets, noted D _i, =1, 2,, and the complement of D _i in D was noted D _-i;

For each D _-i, model i (X) of the first half training XGBoost and model m (X) of the second half training XGBoost are used, respectively, and D _i is used for bayesian optimization to find the best hyper-parameter combination of the models;

calculating an estimated value and a residual error of a substrate shear force Y and a reason variable T corresponding to a sub-dataset D _i by using a trained XGBoost model;

Calculating causal effect corresponding to the sub-data set D _i by using the estimated values of the substrate shearing force Y and the reason variable T corresponding to the sub-data set D _i and residual errors Estimate of (2)ThenThe final causal effect index corresponding to the alternative parameter is obtained.

Further, the screening the first seismic records according to the second tuning parameter to obtain final seismic records for time-course analysis includes:

And counting to obtain the median of the second tuning parameters in the first seismic records, and selecting the seismic records which are different from the median by not more than 25% as final seismic records for time interval analysis.

In a second aspect, the present invention provides a tuning device for seismic records in elastic time-course analysis, comprising:

the first tuning module is used for carrying out amplitude modulation on the initial seismic records adopted in the elastic time course analysis by adopting the effective ground peak acceleration EPA, and screening the initial seismic records according to a preset first tuning parameter to obtain first seismic records;

And the second tuning module is used for filtering the first seismic records according to the second tuning parameters to obtain final seismic records for time-course analysis.

In a third aspect, the present invention provides an electronic device comprising:

At least one processor, and

A memory communicatively coupled to the at least one processor;

wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a method of tuning seismic records in elastic time-course analysis according to the first aspect of the invention.

In a fourth aspect, the present invention provides a non-transitory computer readable storage medium, wherein computer instructions are stored in the storage medium, and when the computer instructions are executed by a computer, the method for tuning seismic records in elastic time-course analysis according to the first aspect of the present invention is implemented.

The method has the advantages that on the basis of adopting the effective ground peak acceleration EPA to perform initial adjustment and selection on the ground vibration record, the causal inference method is adopted to perform adjustment and selection again on the initial adjustment and selection result, so that the discreteness of the shearing force of the final structure substrate can be well controlled, and the correction of the vibration mode decomposition reaction spectrum result by using the elastic time-course analysis result is realized.

Drawings

FIG. 1 is a flow chart of a method for selecting seismic records in an elastic time-course analysis according to an embodiment of the present invention;

FIG. 2 is a finite element model of a building frame structure provided in accordance with an embodiment of the present invention;

FIG. 3 is a causal inference analysis result of seismic parameters according to an embodiment of the invention;

FIGS. 4a to 4d are frequency distribution histograms of main cause variables in the embodiment of the present invention;

FIG. 5 is a schematic diagram of a cross-validation process in accordance with an embodiment of the present invention;

Fig. 6 shows a block diagram of an exemplary electronic device that can be used to implement an embodiment of the invention.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

In the description of the present application, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more of the described features. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.

In the description of the present application, the term "for example" is used to mean "serving as an example, instance, or illustration. Any embodiment described as "for example" in this disclosure is not necessarily to be construed as preferred or advantageous over other embodiments. The following description is presented to enable any person skilled in the art to make and use the application. In the following description, details are set forth for purposes of explanation. It will be apparent to one of ordinary skill in the art that the present application may be practiced without these specific details. In other instances, well-known structures and processes have not been described in detail so as not to obscure the description of the application with unnecessary detail. Thus, the present application is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

As shown in fig. 1, an embodiment of the present invention provides a method for tuning a seismic record in an elastic time-course analysis, including:

S100, amplitude modulation is carried out on the initial earthquake motion record adopted in the elastic time course analysis by adopting the effective ground peak acceleration EPA, and the initial earthquake motion record is screened according to a preset first tuning parameter, so that a first earthquake motion record is obtained.

Specifically, in this embodiment, the initial earthquake motion record used in the elastic time-course analysis is amplitude-modulated by using the effective ground peak acceleration EPA, and the initial earthquake motion record obtained by screening the amplitude-modulated initial earthquake motion record has a difference value of not more than 20% between the average response spectrum of the initial earthquake motion record and the designed response spectrum at the main periodic point of the structure, as the first earthquake motion record. (hereinafter, the method of screening the first seismic record will be referred to as a normalization method).

S200, adopting causal inference to screen out the earthquake motion parameters which directly cause the shearing force of the structural substrate as second tuning parameters.

The method specifically comprises the following steps:

s201, extracting the seismic parameters contained in the first seismic record as candidate parameters of causal inference.

The embodiment of the invention selects common pulse, acceleration and speed related parameters, and takes 17 earthquake motion parameters as alternative parameters. 300 seismic records with fault distances less than 60km are selected from THE PACIFIC Earthquake ENGINEERING RESEARCH (PEER) ground motion database, and corresponding seismic parameters are counted, wherein 90 pulse-type seismic records are included. The 17 alternative vibration parameters and the statistical results are shown in table 1.

TABLE 1 vibration multiple parameters and statistics Table 1Multi-PARAMETERS OF GROUND MOTION AND THE STATISTICS results

Note that the structure has a long axis direction of X and a short axis direction of Y.

S202, carrying out causal inference between the alternative parameters and the structural substrate shear force by adopting a potential result model.

Causal inferences between seismic parameters and structural substrate shear are made using a latent results model (Rubin Causal Model, RCM). RCM assumes that everything except the intervention and the potential outcome is a confounding variable and makes causal inferences by predicting the change in potential outcome before and after the intervention, introducing do operators herein to represent the intervention on the input seismic parameters. For any seismic record, the causal effect index of the alternative parameters on the substrate shear force can be written as,

θ=Y[do(T=1)]- [do(T=0)] (3)

Where T is some intervention on the seismic parameters, t=1 and t=0 denote the intervention/no intervention respectively, Y [ do (t=1) ] denotes the potential substrate shear result from the seismic record under the intervention of the cause variable T, and Y [ do (t=0) ] denotes the potential substrate shear result from the seismic record without the intervention.

The intervention T of the local vibration parameter is a false causative variable of the potential substrate shear result Y, which, in turn,

P[Y|T=t]≠P[Y|do(T=t)] (4)

Wherein, P [ Y|T=t ] is the conditional probability of the substrate shearing force under the intervention T=t, wherein the substrate shearing force is an actual measurement value, and P [ Y|do (T=t) ] is the conditional probability of the substrate shearing force under the intervention T=t, wherein the substrate shearing force is a potential result.

And quantitatively analyzing the causal relation between the alternative parameters and the substrate shearing force under a potential result model frame by using a double-machine learning method to obtain causal effect indexes of the alternative parameters on the substrate shearing force.

The traditional machine learning method can model high-dimensional data and is suitable for solving the causal inference problem of the seismic parameters containing high-dimensional confusion variables, but the traditional machine learning method only pays attention to the prediction effect, ignores possible overfitting, and leads to biased estimation results. According to the existing research, the Double machine learning method (Double MACHINE LEARNING, DML) can effectively avoid over-fitting and eliminate deviation by performing secondary fitting on residual errors generated by the first fitting result, and is more suitable for causal inference. The embodiment of the invention adopts a double-machine learning method to quantitatively analyze the causal relationship between the alternative parameters and the substrate shearing force under the RCM framework. In the case of a sufficient sample being selected, assuming that all candidate parameters are classified into a causal variable T and a confounding variable X, the causal relationship of structural substrate shear to seismic parameters can be expressed as follows,

Y=θT+g(X)+∈ (5)

T=f(X)+η (6)

Wherein Y is the substrate shearing force generated after the intervention of the alternative parameter, theta is the causal effect of the alternative parameter on the substrate shearing force, g (X) and f (X) are unknown functions related to X, and E and eta are both unknown random errors with the mean value of 0.

For the first step of fitting in a double-machine learning method, regularization terms are introduced into XGBoost algorithm (Extreme gradient boosting algorithm), Y and T are fitted respectively, the possibility of overfitting is reduced, meanwhile, in view of the fact that the accuracy of the XGBoost algorithm prediction depends on the choice of super parameters, the optimal super parameters of the XGBoost algorithm are searched by adopting Bayesian optimization technology (Bayesian optimization technique). The principle of regression analysis is to fit the best projection of the dependent variable on the feature space formed by the independent variable, and the residual is perpendicular to the sample space formed by the independent variable, so that the influence caused by the correlation of the independent variable can be eliminated to the greatest extent by fitting the residual. The model obtained by regression analysis is recorded as

For the second step of fitting in the two-machine learning method, linear regression is performed (Linear regression) on the residuals of the XGBoost algorithm fit results, thereby eliminating the estimated bias from the causal variable T and confounding variable X correlations. The residuals corresponding to Y and T may be written as,

By performing linear regression calculation θ with an intercept of 0 on the residuals of Y and T, the simultaneous equations (5), (7), (9) and (10) are obtained,

ε_Y＝θε_T+∈ (11)

Where n is the total amount of data in the designated training data set.

S203, selecting an alternative parameter with the largest causal effect index as a second tuning parameter.

And S300, screening the first earthquake motion record according to the second screening parameters to obtain a final earthquake motion record for time-course analysis.

Specifically, the median of the second tuning parameters in the first seismic records is obtained through statistics, and the seismic records which are different from the median by not more than 25% are selected as final seismic records for time-course analysis.

In addition, the double machine learning method generally needs to perform multiple division to obtain the data set by Cross Validation (Cross Validation)To account for the randomness of the data. Thus, in a preferred embodiment, as shown in fig. 5, step S202 further includes, randomly dividing the first seismic record D into b sub-datasets, denoted as D _i, =1, 2,, and the complement of D _i in D as D _-i, training XGBoost model l (X) and XGBoost model m (X) for each D _-i using the first half and the second half respectively, and using D _i for bayesian optimization to find the best superparameter combination of the models, using the trained XGBoost model to obtain the estimates (equations (7) and (8)) and residuals (equations (9) and (10)) of the base shear Y and the causal variable T corresponding to the sub-dataset D _i, and calculating the causal effect corresponding to the sub-dataset D _i using the estimates and residuals of the base shear Y and the causal variable T corresponding to the sub-dataset D _i Estimate of (2)ThenThe final causal effect index corresponding to the alternative parameter is obtained.

The following describes a method for selecting seismic records in the elastic time-course analysis according to the embodiment of the present invention with reference to specific examples.

A multi-layer framework structure (framework W2) of a museum is chosen as an example, which is severely damaged in a 6.8-level earthquake. The multi-layer building is of a two-layer frame structure, the heights of a first layer and a second layer are respectively 4.8m and 4.2m, and the typical column net span is 7.8m. According to the earthquake region map, the position of the museum belongs to a 9-degree region (0.4 g), the II-type sites are grouped into a third group, and the characteristic period is 0.45s. In this example, the framework W2 is subjected to elastic time-course analysis by using general analysis software ANSYS, and a finite element model of the framework W2 is shown in fig. 2. The first four-order self-vibration periods of the frame W2 are respectively as follows: 0.43s,0.42s,0.36s,0.16s. According to the finite element analysis result and the rules of related specifications, the frame W2 has torsional irregularity, concave-convex irregularity and discontinuous vertical lateral force resisting members, and three irregularity items in total belong to special irregularity structures, and the elastic time interval analysis is needed to be adopted for carrying out supplementary calculation.

According to 300 groups of base shear data of the frame W2, causal inference is performed on the seismic parameters in the table 1 by adopting causal inference, algebraic average values of causal effect indexes of the seismic parameters in two horizontal directions are used as causal effect indexes theta of the seismic parameters, and theta is normalized, and the final analysis result is shown in figure 3. As shown in fig. 3, the cause variables that cause the structural substrate shear force are mainly acceleration-related parameters, and Park-Ang index and Housner strength are the main cause variables that cause the structural substrate shear force. As described above, the fundamental natural vibration period of the frame W2 is less than the field signature period, the seismic response is in the acceleration control section, and the causal inference analysis results are consistent with engineering experience. The frequency distribution histogram of the earthquake motion parameters with higher causal effect index is shown in fig. 4 a-4 d.

According to the method, the seismic records are selected (1) according to a standard method, and the amplitude modulation coefficient S _F is controlled to be less than or equal to 3;

EPA _max is the maximum value of the effective ground peak acceleration, and takes the value according to the regulation of the relevant regulation, EPA _i is the effective ground peak acceleration recorded by the ith ground vibration;

(2) Based on the seismic record selected by the standard method, taking the data with the discrete control parameter value between 25% and 75% quantiles as a final seismic record sample.

The method achieves the aim of reducing the shear discreteness of the substrate by controlling the fluctuation range of the screening parameters, so that the result of elastic time course analysis and inspection can be better used for correcting the design result of the vibration mode decomposition reaction spectrum method. After the substrate shear force discreteness is effectively controlled, the number of earthquake motion records required to be adopted for elastic time course analysis should be correspondingly adjusted. The sample size of the seismic records employed in the time course analysis may be determined according to the following equation,

Where m is the number of seismic records to be used in the elastic time course analysis, Z is related to the confidence level, z=1.96 when the confidence level is 95%, σ is the standard deviation of the overall sample, E is the allowable error, e=35% when three seismic records are used for the time course analysis, e=20% when seven seismic records are used for the time course analysis, and μ is the mean of the overall sample, according to the rules of the related specifications.

120 Pieces of 60 pieces of earthquake motion records meeting the requirements are respectively obtained according to the standard method and the method provided by the embodiment of the invention.

And importing the seismic records subjected to regulation and selection by a standard method into a finite element model of the frame W2 for time-course analysis, and respectively counting the elastic time-course analysis results and the number of the seismic records required to be adopted according to the allowable error of 35% and 20%, wherein the statistical results are shown in Table 2. As shown in Table 2, when the discrete control parameters in the method adopt main cause variables, the discrete property of the substrate shearing force obtained by the elastic time-course analysis can be effectively controlled, and conversely, when the earthquake motion parameters with small causal effect indexes or non-causal variables are selected, the discrete property of the substrate shearing force can not be effectively controlled. When the discrete control parameter selects Housner intensity Pa and Park-Ang index IC, the standard deviation of the substrate shearing force obtained by time course analysis is reduced by 26% and 25% respectively compared with that of the standard deviation method. It is noted that the discretion of the seismic parameters themselves affects the discretion of the final substrate shear result, and because the discretion of the Park-Ang index obtained by selecting the seismic records and the Housner intensity data themselves is large, the standard deviation of the substrate shear calculated by the seismic records selected according to the two indexes is possibly overestimated. After the earthquake motion record is screened for the second time according to the discreteness control parameter, the artificial earthquake motion record can be omitted during the elastic time course analysis. In contrast, when the method is not adopted or the discrete control parameter is not adopted, the main reason variable is not adopted, the discrete nature of most of substrate shear samples is larger, and more earthquake motion records or a certain amount of manual records are needed to be adopted for elastic time-course analysis so as to reduce the discrete nature of the result. According to the causal deducing result and the statistics of the elastic time course analysis result of the seismic records after the adjustment, for the elastic time course calculation of the frame W2 in the museum, the effective peak acceleration EPA and Park-Ang index IC are suggested to be adopted as the adjustment parameters of the seismic records.

TABLE 2 elastic time course analysis results statistics Table 2The statistics of elastic time-history analysis results of seismic records after tuning

Note that the discretionary control parameter data in the method uses structure primary direction data.

The embodiment of the invention also provides a device for adjusting and selecting the earthquake motion record in the elastic time course analysis, which comprises the following steps:

According to one aspect of the disclosure, an electronic device is also disclosed that includes at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executable by the at least one processor to enable the at least one processor to perform the method described above.

According to one aspect of the present disclosure, a non-transitory computer-readable storage medium having stored therein computer instructions that when executed by a computer implement the above-described method is also disclosed.

According to one aspect of the present disclosure, a computer program product is also disclosed, comprising a computer program, wherein,

The computer program, when executed by a processor, implements the above-described method.

Referring to fig. 6, a block diagram of an electronic device 600 that may be a server or a client of the present disclosure, which is an example of a hardware device that may be applied to aspects of the present disclosure, will now be described. Electronic devices are intended to represent various forms of digital electronic computer devices, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 6, the electronic device 600 includes a computing unit 601 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 602 or a computer program loaded from a storage unit 608 into a Random Access Memory (RAM) 603. In the RAM603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, ROM 602, and RAM603 are connected to each other by a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.

A number of components in the electronic device 600 are connected to the I/O interface 605, including an input unit 606, an output unit 607, a storage unit 608, and a communication unit 609. The input unit 606 may be any type of device capable of inputting information to the electronic device 600, the input unit 606 may receive input numeric or character information and generate key signal inputs related to user settings and/or function control of the electronic device, and may include, but is not limited to, a mouse, a keyboard, a touch screen, a trackpad, a trackball, a joystick, a microphone, and/or a remote control. The output unit 607 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, video/audio output terminals, vibrators, and/or printers. Storage unit 608 may include, but is not limited to, magnetic disks, optical disks. The communication unit 609 allows the electronic device 600 to exchange information/data with other devices through a computer network, such as the internet, and/or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers and/or chipsets, such as bluetooth (TM) devices, 802.11 devices, wiFi devices, wiMax devices, cellular communication devices, and/or the like.

The computing unit 601 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 601 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the respective methods and processes described above, for example, an image processing method. For example, in some embodiments, the image processing method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 600 via the ROM 602 and/or the communication unit 609. When a computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the image processing method described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform the image processing method by any other suitable means (e.g. by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include being implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be a special or general purpose programmable processor, operable to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user, for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback), and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a Local Area Network (LAN), a Wide Area Network (WAN), and the Internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel, sequentially or in a different order, provided that the desired results of the disclosed aspects are achieved, and are not limited herein.

Although embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it is to be understood that the foregoing methods, systems, and apparatus are merely exemplary embodiments or examples, and that the scope of the present invention is not limited by these embodiments or examples but only by the claims following the grant and their equivalents. Various elements of the embodiments or examples may be omitted or replaced with equivalent elements thereof. Furthermore, the steps may be performed in a different order than described in the present disclosure. Further, various elements of the embodiments or examples may be combined in various ways. It is important that as technology evolves, many of the elements described herein may be replaced by equivalent elements that appear after the disclosure.

Claims

1. A method for selecting earthquake records in elastic time history analysis, comprising:

The effective ground peak acceleration EPA is used to amplitude modulate the initial seismic motion record used in the elastic time history analysis, and the initial seismic motion record is screened according to a preset first selection parameter to obtain a first seismic motion record;

Causal inference is used to select the ground motion parameters that directly cause the shear force of the structural base as the second adjustment parameter;

The first seismic record is screened according to the second selection parameter to obtain a final seismic record for time history analysis.

2. The selection method according to claim 1 is characterized in that the initial seismic motion record is screened according to a preset first selection parameter to obtain the first seismic motion record, including: screening the initial seismic motion record after amplitude modulation, the difference between the average response spectrum of the initial seismic motion record at the main period point of the structure and the design response spectrum is not more than 20% as the first seismic motion record.

3. The selection method according to claim 1 is characterized in that the use of causal inference to select the ground motion parameters that directly cause the shear force of the structural base as the second selection parameters includes:

extracting earthquake motion parameters contained in the first earthquake motion record as candidate parameters for causal inference;

Using a potential result model to make causal inferences between the alternative parameters and the structural base shear;

The dual machine learning method is used to quantitatively analyze the causal relationship between the candidate parameters and the base shear in the framework of the potential result model, and the causal effect index of each candidate parameter on the base shear is obtained;

The alternative parameter with the largest causal effect indicator is selected as the second adjustment parameter.

4. The selection method according to claim 3 is characterized in that the use of a potential result model to perform causal inference between the candidate parameters and the structural base shear force comprises:

Assuming that the alternative parameters include two types: cause variable T and confounding variable X, the causal relationship between the structural base shear and the alternative parameters is expressed as:

Y = θT + g(X) + ∈;

T = f(X) + η;

where Y is the base shear generated after the intervention of the alternative parameters; g(X) and f(X) are unknown functions of X; ∈ and η are both unknown random errors with mean 0; θ is the causal effect of the alternative parameters on the base shear, θ = Y[do(T = 1)] - Y[do(T = 0)], T = 1 and T = 0 represent intervention/no intervention, respectively; Y[do(T = 1)] represents the potential base shear result generated by the seismic record under the intervention of the causal variable T; Y[do(T = 0)] represents the potential base shear result generated by the seismic record without intervention.

5. The selection method according to claim 4 is characterized in that the use of a dual machine learning method to quantitatively analyze the causal relationship between the candidate parameters and the base shear under the framework of the potential result model to obtain the causal effect index of each candidate parameter on the base shear includes:

The XGBoost algorithm is used to fit the base shear force Y and the cause variable T generated after the intervention of the alternative parameters, and the Bayesian optimization algorithm is used to search for the optimal hyperparameters of the XGBoost algorithm to obtain the estimated values of the base shear force Y and the cause variable T. and

Compute the fitted residuals:

Perform a linear regression with an intercept of 0 on the fitted residuals to calculate the estimated value of the causal effect θ of the alternative parameters on the base shear

ε _Y =θε _T +∈;

Where n is the total amount of data in the specified training data set.

6. The selection method according to claim 5 is characterized in that the use of a dual machine learning method to quantitatively analyze the causal relationship between the candidate parameters and the base shear under the framework of the potential result model to obtain the causal effect index of each candidate parameter on the base shear also includes:

The first ground motion record D is randomly divided into b sub-data sets, denoted as D _i , i=1,2,…,b, and the complement of D _i in D is denoted as D _-i ;

For each D _-i , use the first half to train the XGBoost model l(X) and the second half to train the XGBoost model m(X), and use D _i to perform Bayesian optimization to find the best hyperparameter combination of the model;

The trained XGBoost model is used to obtain the estimated values and residuals of the base shear force Y and the cause variable T corresponding to the sub-dataset D _i ;

Using the estimated values and residuals of the base shear Y and the cause variable T corresponding to the sub-dataset D _i , the causal effect corresponding to the sub-dataset D _i is calculated Estimated value of but That is the final causal effect indicator corresponding to the alternative parameter.

7. The selection method according to claim 1, wherein the step of screening the first seismic record according to the second selection parameter to obtain the final seismic record for time history analysis comprises:

The median of the second selected parameter in the first seismic record is obtained by statistics, and the seismic record whose difference with the median is not greater than 25% is selected as the final seismic record for time history analysis.

8. A device for selecting earthquake records in elastic time history analysis, characterized by comprising:

The first selection module uses the effective ground peak acceleration EPA to modulate the initial seismic motion record used in the elastic time history analysis, and selects the initial seismic motion record according to the preset first selection parameter to obtain the first seismic motion record;

The second selection module uses causal inference to select the seismic motion parameters that directly cause the shear force of the structural base as the second selection parameters; the first seismic motion record is selected according to the second selection parameters to obtain the final seismic motion record for time history analysis.

9. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor;

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for selecting seismic records in elastic time history analysis as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium, characterized in that the storage medium stores computer instructions, and when the computer instructions are executed by a computer, the method for selecting seismic records in elastic time history analysis according to any one of claims 1 to 7 is implemented.