Seismic Parameters Research Articles

Evaluation of liquefaction potential in central Taiwan using random forest method

AbstractLiquefaction is a significant geotechnical hazard in seismically active regions like Taiwan, threatening infrastructure and public safety. Accurate prediction models are essential for assessing soil susceptibility to liquefaction during seismic events. This study evaluates liquefaction potential in central Taiwan using the random forest (RF) method. The RF models were developed with a dataset of 540 soil and seismic parameter sets, including depth, effective and total overburden stresses, SPT-N values, fine soil content, earthquake magnitude, peak ground acceleration, and historical liquefaction occurrences. Rigorous validation techniques, such as cross-validation and comparisons with observed liquefaction events, confirm the RF model’s effectiveness, achieving an accuracy of 98.89%. The model also quantifies predictor importance, revealing that the SPT-N value is the most critical soil factor, while peak ground acceleration is the key seismic factor for liquefaction prediction. Notably, the RF model outperforms simplified procedures in accuracy, even with fewer input factors. Our case studies show that an accuracy of over 95% can still be achieved, highlighting the RF model’s superior performance compared to conventional methods, which struggle to reach similar levels.

A Restoring Force Correction Model for Enveloped Steel Jacket-Confined Seismic-Damaged Rectangular Recycled Aggregate Concrete-Filled Steel Tubular Columns

In order to establish a suitable restoring force correction model for enveloped steel jacket (ESJ)-confined seismic-damaged rectangular recycled aggregate concrete-filled steel tubular (RRACFST) columns, based on experimental research, a study of the seismic performance and parameters of ESJ-confined seismic-damaged RRACFST columns was carried out. The restoring force theory, model test, and OpenSees simulation of ESJ-confined seismic-damaged RRACFST columns were conducted. Firstly, a trilinear model of the skeleton curve and a suitable restoring force model for ESJ-confined seismic-damaged RRACFST columns were established. The results were compared with the model test results, and it was found that the two results had good consistency. Secondly, the initial damage of the RRACFST column was simulated by the reducing material properties method, and a correct numerical model for ESJ-confined seismic-damaged RRACFST columns was proposed. The influence mechanism of seismic parameters of the RRACFST column was clarified. Finally, the seismic parameter combination with the best seismic performance for ESJ-confined seismic-damaged RRACFST columns was established; namely, the replacement rate of recycled coarse aggregate is 50%, the concrete strength is C40, the axial compression ratio is 0.3, the strength of the rectangular steel tube is Q345, the wall thickness of the steel tube is 4 mm, and the slenderness ratio is 7.5.

Large-scale geohazards risk of submarine landslides considering the subsea cables vulnerability: A case study from the northern continental slopes of South China Sea

Submarine landslides pose significant threats to subsea cables distributed on the global seabed. However, regional scale risk assessment of landslide geohazards is rarely reported. This study introduces a methodology for regional-scale geohazard risk prediction of submarine landslides, focusing on the northern continental slopes of the South China Sea. Initially, the study employed the infinite-slope method to calculate safety factors for typical submarine slopes. Addressing uncertainties in geotechnical and seismic parameters, Monte Carlo simulations determine slope failure probabilities. Using kriging interpolation, localized failure probabilities are extrapolated to regional scales, establishing a spatiotemporal distribution method for large-scale geohazard susceptibility. Hazard levels are subsequently determined considering the volumes of potential landslides. The density of subsea cables is used as a vulnerability factor to guide the regional-scale assessment of cable vulnerability. Finally, integrating vulnerability with hazard levels provides a comprehensive assessment of landslide-induced large-scale geohazard risks. The findings highlight elevated geohazard risks in the Taiwan Bank slope segment, moderate risks in the Zhujiang Valley and Shenhu slope segments, with lower risks in other areas.

Seismic Collapse Risk Assessment and Fragility Analysis of Reinforced Masonry Core Walls with Boundary Elements Using the FEMA P695 Methodology

In the past, the primary objective of structural design codes was centred around ensuring human life safety, with a strong focus on preventing structural collapse. This approach predominantly relied on force- or strength-based criteria as the basis for design. Nevertheless, a notable shift has occurred recently, transitioning the design philosophy from 'strength' towards a 'performance'-based approach. This shift signifies a growing recognition that strength and performance are not identical. However, to adopt the performance-based seismic design approach, it is essential to develop precise models for estimating damage and potential losses associated with various seismic force-resisting systems (SFRSs). Fragility functions are widely interpreted among the most prevalent damage and loss models. They establish a crucial link between specific demand parameters and the likelihood of exceeding various damage states. Although reinforced masonry (RM) structures have recently gained popularity, the seismic design of mid- to high-rise RM structures is still challenging because it requires a reliable SFRS capable of providing the needed ductility and capacity. Therefore, the main objective of this study is to evaluate the key seismic performance parameters (i.e., system overstrength and ductility) and to perform a collapse capacity risk assessment of reinforced masonry core walls with boundary elements (RMCW+BEs) as the main SFRS in RM structures. The current study utilizes the applied element method (AEM) implemented in the Extreme Loading for Structures software (ELS) to model three RM structures with various heights (10-, 15-, and 20-story buildings). Nonlinear pseudo- static pushover analysis was performed to quantify the ductility and overstrength of the proposed RM system following the FEMA P695 guidelines. In addition, an incremental dynamic analysis (IDA) was carried out to assess the seismic collapse risk of RMCW+BEs by generating system-level-based fragility curves. The results showed that the proposed RM system provides the needed ductility, overstrength and deformation capacity for a ductile SFRS for typical mid- and high-rise RM buildings. Furthermore, the developed collapse fragility curves verified excellent seismic performance, negligible collapse probability values and high reserved collapse capacity at the maximum considered earthquake (MCE) design level. The findings of this study significantly contribute toward adopting RMCW+BEs as an effective SFRS for typical RM buildings in the next generations of North American masonry design standards.

Seismic Collapse Risk Assessment and Fragility Analysis of Reinforced Masonry Core Walls with Boundary Elements Using the FEMA P695 Methodology

In the past, the primary objective of structural design codes was centred around ensuring human life safety, with a strong focus on preventing structural collapse. This approach predominantly relied on force- or strength-based criteria as the basis for design. Nevertheless, a notable shift has occurred recently, transitioning the design philosophy from 'strength' towards a 'performance'-based approach. This shift signifies a growing recognition that strength and performance are not identical. However, to adopt the performance-based seismic design approach, it is essential to develop precise models for estimating damage and potential losses associated with various seismic force-resisting systems (SFRSs). Fragility functions are widely interpreted among the most prevalent damage and loss models. They establish a crucial link between specific demand parameters and the likelihood of exceeding various damage states. Although reinforced masonry (RM) structures have recently gained popularity, the seismic design of mid- to high-rise RM structures is still challenging because it requires a reliable SFRS capable of providing the needed ductility and capacity. Therefore, the main objective of this study is to evaluate the key seismic performance parameters (i.e., system overstrength and ductility) and to perform a collapse capacity risk assessment of reinforced masonry core walls with boundary elements (RMCW+BEs) as the main SFRS in RM structures. The current study utilizes the applied element method (AEM) implemented in the Extreme Loading for Structures software (ELS) to model three RM structures with various heights (10-, 15-, and 20-story buildings). Nonlinear pseudo- static pushover analysis was performed to quantify the ductility and overstrength of the proposed RM system following the FEMA P695 guidelines. In addition, an incremental dynamic analysis (IDA) was carried out to assess the seismic collapse risk of RMCW+BEs by generating system-level-based fragility curves. The results showed that the proposed RM system provides the needed ductility, overstrength and deformation capacity for a ductile SFRS for typical mid- and high-rise RM buildings. Furthermore, the developed collapse fragility curves verified excellent seismic performance, negligible collapse probability values and high reserved collapse capacity at the maximum considered earthquake (MCE) design level. The findings of this study significantly contribute toward adopting RMCW+BEs as an effective SFRS for typical RM buildings in the next generations of North American masonry design standards.

MODIFICATION OF THE ATTENUATION EQUATION FOR PEAK GROUND ACCELERATION (PGA) IN THE NORTH SUMATERA REGION

Prediction of the empirical formula for ground acceleration is an important thing to analyze for seismic estimation in an area. This needs to be done to reduce the negative impact of earthquakes as a more appropriate mitigation effort in planning and designing earthquake-resistant buildings. Therefore, the empirical formulation from Zhaou, et al is used, which is well adapted to the seismotectonic conditions of the North Sumatra region, which has high seismic vulnerability, considering that this region is located in an active subduction zone. This advantage allows the model to more accurately predict the maximum ground acceleration (PGA) produced by earthquakes in the region. The purpose of this research is to obtain an empirical formulation of the maximum ground acceleration value based on seismic parameters such as distance, depth, and magnitude. It will also show the relationship between these parameters and the value of maximum ground acceleration or peak ground acceleration (PGA) in the North Sumatra region. This study modifies the empirical formulation of Zhao, et al data sourced from the Meteorology Climatology and Geophysics Agency (BMKG) of North Sumatra in 2017-2023 with a magnitude of 3 - 6 Mw. This research uses non-linear regression with the least squares method. The results of the analysis of the empirical formula produce the constant value sought a = 1.4703 b = -0.0025, c = 20.7441, d = 0.0196, e = 0.0015, S_SS = -2.0843, and S_SL = -0.0529. An empirical formula was obtained for the North Sumatra region based on the research results. This equation can be used on a scale of 3.0 - 6.0 Mw and a distance to the earthquake source between 0 - 300 km. The relationship between each parameter of this empirical formula is that the PGA value will increase with the magnitude, and the PGA value will decrease as the epicentre distance increases.

Analysis of Soil-Structure Interaction of Framed Structure

This study examines the soil-structure interaction (SSI) effects on the seismic behavior of G+7 and G+12 reinforced concrete buildings. Using finite element modelling in SAP2000, we assess key seismic parameters, including base shear, time period, lateral displacement, and footing settlement, across three soil types: soft, medium, and hard. This comparative analysis provides insights into how building height and soil flexibility influence structural stability, suggesting optimal design considerations for enhancing earthquake resilience. Key Words: Soil-Structure Interaction (SSI), Base Shear, Lateral Displacement, Seismic Analysis, G+7, G+12, SAP2000, Soft Soil, Medium Soil, Hard Soil.

Dynamics and Detection of Pulsed Tremor at Whakaari (White Island), Aotearoa New Zealand

AbstractVolcanic tremor is a crucial indicator for assessing the state and hazard potential of volcanic systems. At Whakaari (White Island volcano, Aotearoa New Zealand), a pulsed tremor signal emerged after a hydrothermal explosion in August 2012. The tremor accompanied the extrusion of a lava dome, before gradually disappearing prior to the onset of renewed hydrothermal activity in January 2013. We interpret this seismic signal to represent discrete gas transfers from a magmatic intrusion toward a permeable cap—possibly a hydrothermal seal—in the upper layers of Whakaari's hydrothermal system. Such tremor may thus be associated with heightened potential for hazardous explosive activity but is difficult to detect using conventional seismic monitoring parameters. To highlight the emergence of subtle periodic signals, we experiment with Lomb‐Scargle periodograms (LS). LS detect the tremor 5 days before it becomes visible in seismograms, thus facilitating the recognition of such elusive seismic patterns.

Parameters Affecting Seismic Performance of Longitudinally Connected Track Lateral Blocks on Bridges Under Low-Cycle Reciprocating Loads

In the current seismic analysis of high-speed railway track–bridge structures, the constitutive parameters of track lateral blocks (TLBs) are not uniform. A finite element model of the TLB’s main beam is established, which considers concrete plastic damage, interface bond between old and new concrete, and bond slip between anchorage steel bars (ASBs) and concrete. Quasi-static tests of nine TLBs with different numbers and ASB diameters at a 1:2 scale was carried out to verify the accuracy of the TLB finite element model in terms of the failure pattern, interface damage, hysteretic properties, and skeleton model. Based on the verified TLB finite element model, the influence of different TLB design parameters on its seismic performance parameters was studied. The results show that the established TLB finite element model is in good agreement with the test data in terms of failure pattern, interface damage, hysteretic properties, and skeleton model, which verifies the accuracy of the TLB finite element model. ASB number has the greatest influence on the TLB yield load, followed by ASB diameter, concrete strength, ASB strength, and ASB spacing. ASB diameter has the greatest influence on the TLB peak load, followed by ASB number, ASB strength, concrete strength, and ASB spacing. ASB diameter has the greatest influence on the TLB energy dissipation capacity, followed by ASB number, concrete strength, ASB strength, and ASB spacing. With the increase in concrete strength and ASB strength, TLB ductility decreases. When the ASB number is higher than eight or the diameter is higher than 12 mm and the corresponding interfacial steel bar ratio reaches 0.82%, the TLB ductility will decrease.

Seismic Discrimination Using Regional Distance Coda-Wave Ratios

Abstract This study utilizes regional seismic coda-wave amplitude ratios to distinguish earthquakes and declared nuclear explosions in the Democratic People’s Republic of Korea (DPRK). We compute the coda spectral ratios at common stations to isolate source information from propagation and site effects. The resulting frequency-dependent amplitude ratios help identify differences between source types at DPRK. We use parametric models to validate observations, linking them to seismic source parameters. Our method effectively distinguishes the spectral signatures of explosions, nearby earthquakes, and a post-explosion collapse event. The measurements show clear differences between source-type combinations and agree with estimated seismic source models. Notably, the coda spectral ratios between explosions and earthquakes indicate that explosive sources exhibit relatively lower high-frequency S-wave energy and corner frequency compared to P waves. We leverage these spectral differences to develop a Bayesian inference framework for seismic source discrimination at the DPRK test site. This study demonstrates that spectral differences between source types can be exploited in coda spectral ratios to provide critical information for seismic discrimination.

K-means for earthquakes: disaggregation analyses of small events by considering wave components and soil types

In this paper, k-means algorithm has been used to disaggregate seismic parameters to evaluate their inter-correlations. A goal is to quantify in a disaggregated way the weights and effects of each parameter with respect to other ones. From the database, about 4900.0 data, divided into 22.0 categories, have been collected. The main divisions regard the wave components in horizontal and vertical axis and the soil characteristics. The studied seismic zone is the “Norpirenaica oriental,” placed in the Pyrenees area between Spain and France, classified as a very high seismic hazard. Numerical and analytical analyses have been carried out to implement the algorithm. Preliminary analyses and results would quantify the role of the sand horizontal stratigraphy, the non-linear effects, the elasticity of the soil, and the energy damping phenomenon. Curves are plotted in stochastic distributions and elastic spectra accelerations. Results show good prediction for vertical spectral accelerations and for far and relative strong events. Rigorously, results are valid only for the studied seismogenic zone under predefined constrictions and ranges.

Monthly Maximum Magnitude Prediction in the North–South Seismic Belt of China Based on Deep Learning

The North–South Seismic Belt is one of the major regions in China where strong earthquakes frequently occur. Predicting the monthly maximum magnitude is of significant importance for proactive seismic hazard defense. This paper uses seismic catalog data from the North–South Seismic Belt since 1970 to calculate and extract multiple seismic parameters. The monthly maximum magnitude is processed using Variational Mode Decomposition (VMD) with sample segmentation to avoid information leakage. The decomposed multiple modal data and seismic parameters together form a new dataset. Based on these datasets, this paper employs four deep learning models and four time windows to predict the monthly maximum magnitude, using prediction accuracy (PA), False Alarm Rate (FAR), and Missed Alarm Rate (MR) as evaluation metrics. It is found that a time window of 12 generally yields better prediction results, with the PA for Ms 5.0–6.0 earthquakes reaching 77.27% and for earthquakes above Ms 6.0 reaching 12.5%. Compared to data not decomposed using VMD, traditional error metrics show only a slight improvement, but the model can better predict short-term trends in magnitude changes.

Study on the attenuation relationship of the acceleration envelope parameters for the Wenchuan earthquake aftershocks

The seismic attenuation relationship between ground motion parameters (such as peak acceleration and response spectra value) and seismic parameters (such as magnitude and epicentral distance) is an important foundation for seismic hazard analysis and the core of determining seismic input parameters for seismic resistance engineering. The acceleration envelope parameters, which describe the relationship between ground motion intensity and time variation, are primarily used for artificially synthesizing seismic motion. At present, little research has been performed on the attenuation relationship of acceleration envelope parameters in the Longmenshan fault zone on the eastern side of the Tibetan Plateau in China. Therefore, this study selected the Mw4-6 aftershock records of the 2008 Wenchuan earthquake that occurred on the Longmenshan fault zone and established a commonly used three segment envelope model parameter attenuation relationship. We classified aftershock records based on their source mechanisms and obtained attenuation relationship models for thrust slip aftershocks, thrust and strike slip aftershocks, and strike slip aftershocks. The results are as follows: (1) The thrust slip aftershock records had the longest rising stage which is the time difference from the arrival of P-waves to the beginning of the stable sustained stage of seismic motion. Thrust and strike slip aftershocks records had the longest stable sustained stage period and the slowest attenuation at the tail of the record. (2) The attenuation relationship of the acceleration envelope parameters commonly used in Chinese engineering for artificially synthesizing seismic motion is based on the strong earthquake records in the western United States. But, compared to the attenuation characteristics of the acceleration envelope function in the western United States, the Mw4-6 earthquake records on the Longmenshan fault zone had a slower attenuation rate at the tail of the record. So, accurately artificially synthesizing seismic motion through envelope parameter attenuation model requires the use of attenuation model established by earthquake records in this region.

Optimization of seismic insulation parameters of frame buildings using adaptive methods

Objective. The purpose of the work is to calculate frame buildings with flexible connections and with flexible and switch-off connections, to make recommendations for the use of seismic insulation methods depending on the characteristics of seismic events. Method. Development of an algorithm for calculating frame buildings of a frame structural system with seismic protection in buildings with flexible connections and with flexible and switchable connections Result. A method for calculating frame buildings of a frame structural system with seismic insulation in the form of flexible connections and flexible and switchable connections has been developed. A computer calculation program for seismically insulated and nonseismically insulated frame buildings in the Delphi 5.0 programming environment in the Object Pascal language using the Runge-Kutta numerical method of the 4th order for solving differential equations has been compiled Conclusion. It is determined that the construction of frame buildings with the seismic insulation methods considered in this paper, namely with flexible connections and with flexible connections and disconnecting connections, is not impractical for territories where long-period fluctuations are possible, due to the fact that excessively high deformations occur.

Long-period ground motion simulation and pulse probability distribution characteristics of the Kumamoto MW 7.1 earthquake in Japan

The 2016 Kumamoto MW 7.1 earthquake in Japan caused severe structural damage. It produced valuable seismic records that offered an opportunity to investigate the characteristics of long-period ground motion and pulse probability distribution. This study works in three main parts. First, we use the finite difference method (FDM) to reproduce the long-period ground motions of the 2016 Kumamoto MW 7.1 earthquake. That provides a crucial foundation for the subsequent phases. We identify velocity pulses within both synthetic and observed waveforms. By rigorously analyzing these pulses, we unveiled a wealth of insights into the characteristics of velocity pulses, strengthening the influence of the rupture directivity effect. In particular, the northeast region of the rupture exhibited a significant concentration of pulsed attributes. Finally, we further focus on establishing statistical relationships between different pulse parameters. The study revealed that the pulse probability distribution model fit well with the observed pulse distribution in the parallel to the fault (FP) and normal to the fault (FN) directions. However, the model also faced challenges in the FN direction due to the mechanism of velocity pulses and altered rupture propagation direction. Crucially, the study established that velocity pulses primarily cluster within a fault distance of less than 30 km, and the relationship between pulse peak value and fault distance followed an exponential trend. The study underscores the imperative of further refining the probability distribution model of velocity pulses, making it more robust and widely applicable. Furthermore, the correlation between velocity pulses and seismic parameters requires further validation through seismic records. This study sets the stage for a deeper understanding of earthquake-induced ground motion and pulse characteristics, contributing to the ongoing efforts in seismic risk assessment.

Mainshock–aftershock sequence simulation via latent space encoding of generative adversarial networks

AbstractAftershocks (ASs) following strong mainshocks (MSs) can exacerbate structural damage or lead to collapse. However, the scarcity of recorded data necessitates reliance on artificial sequences, which have difficulty in characterizing the time‐frequency correlation between MSs and ASs. This study innovatively converts the AS time history prediction into an image translation task, exploiting the invertible transformation between accelerograms and time‐frequency representations. An encoder–decoder neural network is developed to encode the MS information into the latent space of a pre‐trained generative adversarial network, enabling accurate AS predictions through the decoder. The integration of seismic parameters further improves the AS prediction performance. Comparative analyses demonstrate that the proposed method outperforms the traditional ones on accuracy and robustness and reproduces the non‐stationarity of ASs.

Vulnerability-based seismic resilience and post-earthquake recovery assessment for substation systems

The function of substation systems is significantly influenced by earthquakes. To assess the seismic resistance of substation systems and improve post-earthquake recovery efficiency, this study constructs a multi-module framework for seismic vulnerability assessment of substations. Based on the equipment functional status analysis module, a functional network model was established considering the structural function and electrical load of substation systems. The system functional state matrix was obtained through Monte Carlo simulation, and the seismic vulnerability was assessed from the aspects of electrical transmission reliability and the importance of power users. Subsequently, a substation functional index was established based on seismic vulnerability parameters. Post-earthquake functional recovery strategy analysis frameworks were constructed to quantitatively evaluate the recovery process. A stepped functional time-varying function was proposed through iterative analysis, thereby converting the probability parameters of functional status into seismic resilience deterministic indices. Through seismic resilience analysis of a typical 220 kV step-down substation, seismic vulnerability curves for both electrical transmission and power users were obtained. Furthermore, the seismic sensitive intervals of the substation and key equipment for post-earthquake recovery were clarified. Notably, the seismic resilience level and optimal recovery strategy of the typical 220 kV step-down substation were determined.

Declustering characteristics of the North China Plain seismic belt and its effect on probabilistic seismic hazard analysis

The accurate and impartial identification of background seismic activity is an important foundation for earthquake model construction and probabilistic seismic hazard analysis (PSHA). We use the Gardner and Knopoff, Reasenberg, nearest-neighbor and stochastic declustering methods to decluster in the North China Plain seismic belt, analyzing the effect of declustering. We optimize the input parameters of the algorithms through Kolmogorov-Smirnov Poisson distribution tests (e.g., the clustering threshold of the nearest-neighbor method is set to 0.6). Four algorithms demonstrate good declustering effects, removing the impact of the strong seismicity in Tangshan while retaining similar trends to all events. The occurrence of major earthquake events can significantly impact the declustering characteristics in terms of time, space and magnitude. There is a difference in the overall hazard between the four declustering algorithms and the Fifth Generation Seismic Ground Motion Parameters Zonation Map of China. The difference in seismic hazard curves is mainly influenced by the annual average occurrence rate (background rates) and the Gutenberg-Richter b value, with the effect of the b value being more pronounced. The analysis of the effect of algorithms and their impact on PSHA can provide reference for earthquake risk assessment, engineering seismic design and disaster research.

Neural-Network and Multivariate-Normal-Distribution Hybrid Method for Real-Time Ground-Shaking Reconstruction

ABSTRACT Ground-shaking maps provide a spatial representation of the impact of a seismic event in terms of ground-motion parameters (GMPs), especially useful in the context of seismic monitoring and civil protection operations. Algorithms used to compute these maps usually rely on seismic source parameters to steer the interpolation process and consequently are limited to operate in near-real time. The present work introduces a novel algorithm that combines neural networks with the multivariate normal distribution method to reconstruct ground-shaking maps using only data available in real time, improving on previously proposed algorithms. The core idea of the proposed algorithm is to maintain the structure proposed by ShakeMap while removing the dependence on the source parameters, imposed by the use of ground-motion prediction equations, by replacing them with an appropriate neural network working on the GMPs recorded in real time at the seismic stations. The overall workflow of the method and the details of the neural network architecture and training are described. A model trained on synthetic and recorded data to target seismic events affecting the Italian territory is tested using the 2016 Norcia, Italy, earthquake showing the method reconstruction capabilities, its robustness to noise and to network geometry changes, and its real-time potential.

Seismic response prediction method of train-bridge coupled system based on convolutional neural network-bidirectional long short-term memory-attention modeling

Seismic response prediction is crucial for the safety analysis of train-bridge coupled systems. However, due to the complexity, suddenness, and high-risk nature of earthquakes, there are strong nonlinear relationships among different parts of bridges, making it challenging to express their spatial correlations using analytical models and traditional neural networks. To address this, this paper establishes a ballast track shaker scaling model and employs the grating monitoring measurement method to construct a spatial quasi-distributed monitoring system for the ballast track, thereby collecting seismic strain responses of the train-bridge coupled system under various seismic conditions. A hybrid neural network method is proposed for predicting the seismic responses of the train-bridge coupled system. This hybrid neural network integrates the features of a Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory Neural Network (BiLSTM), and the attention mechanism, thereby termed the CNN-BiLSTM-attention hybrid neural network. The model was validated using strain responses from 54 seismic scenarios. The results indicate that the model has a Mean Absolute Error (MAE) of 0.2349 and a coefficient of determination (R2) of 0.9446. Comparing the prediction results with those from RNN and LSTM models, it was found that the CNN effectively extracts features under various seismic parameters, while the BiLSTM better captures the temporal information of the strain responses, ensuring effective prediction regardless of the magnitude of strain responses. Therefore, the CNN-BiLSTM-attention hybrid neural network model is recommended for predicting seismic response.