Prediction of Cumulative Absolute Velocity Based on Refined Second-order Deep Neural Network

Duofa Ji; Jin Liu; Weiping Wen; Changhai Zhai; Wei Wang; Evangelos I. Katsanos

doi:10.1080/13632469.2021.1985017

Prediction of Cumulative Absolute Velocity Based on Refined Second-order Deep Neural Network

Duofa Ji, Jin Liu, Weiping Wen^*, Changhai Zhai, Wei Wang, Evangelos I. Katsanos^*

^*Corresponding author for this work

Department of Civil and Mechanical Engineering

Harbin Institute of Technology

Research output: Contribution to journal › Journal article › Research › peer-review

Abstract

This study aims to develop a reliable ground motion model (GMM) for CAV by using ground motion (GM) recordings from the PEER NGA-West2 database. A total of 17,684 GM recordings are chosen and randomly separated into the training, validation, and testing datasets. The DNN is advanced by incorporating the refined second-order (RSO) neuron. The effect of seismological and site-specific parameters on the predicted CAV is investigated. The comparative assessment of four existing models with the RSO-DNN model of this study highlights the superior prediction skill of the latter one since the RSO-DNN model is found to be associated with considerably less error.

Original language	English
Journal	Journal of Earthquake Engineering
Volume	26
Issue number	15
Pages (from-to)	8021-8040
ISSN	1363-2469
DOIs	https://doi.org/10.1080/13632469.2021.1985017
Publication status	Published - 2022

Keywords

Cumulative absolute velocity
Deep neural network
Ground motion model
PEER NGA-West2 database
Standard deviation

Access to Document

10.1080/13632469.2021.1985017

OpenUrl availability

Full text

Cite this

@article{3a4378ffda834497aaf067dff92e6c73,

title = "Prediction of Cumulative Absolute Velocity Based on Refined Second-order Deep Neural Network",

abstract = "This study aims to develop a reliable ground motion model (GMM) for CAV by using ground motion (GM) recordings from the PEER NGA-West2 database. A total of 17,684 GM recordings are chosen and randomly separated into the training, validation, and testing datasets. The DNN is advanced by incorporating the refined second-order (RSO) neuron. The effect of seismological and site-specific parameters on the predicted CAV is investigated. The comparative assessment of four existing models with the RSO-DNN model of this study highlights the superior prediction skill of the latter one since the RSO-DNN model is found to be associated with considerably less error.",

keywords = "Cumulative absolute velocity, Deep neural network, Ground motion model, PEER NGA-West2 database, Standard deviation",

author = "Duofa Ji and Jin Liu and Weiping Wen and Changhai Zhai and Wei Wang and Katsanos, {Evangelos I.}",

year = "2022",

doi = "10.1080/13632469.2021.1985017",

language = "English",

volume = "26",

pages = "8021--8040",

journal = "Journal of Earthquake Engineering",

issn = "1363-2469",

publisher = "CRC Press/Balkema",

number = "15",

}

TY - JOUR

T1 - Prediction of Cumulative Absolute Velocity Based on Refined Second-order Deep Neural Network

AU - Ji, Duofa

AU - Liu, Jin

AU - Wen, Weiping

AU - Zhai, Changhai

AU - Wang, Wei

AU - Katsanos, Evangelos I.

PY - 2022

Y1 - 2022

N2 - This study aims to develop a reliable ground motion model (GMM) for CAV by using ground motion (GM) recordings from the PEER NGA-West2 database. A total of 17,684 GM recordings are chosen and randomly separated into the training, validation, and testing datasets. The DNN is advanced by incorporating the refined second-order (RSO) neuron. The effect of seismological and site-specific parameters on the predicted CAV is investigated. The comparative assessment of four existing models with the RSO-DNN model of this study highlights the superior prediction skill of the latter one since the RSO-DNN model is found to be associated with considerably less error.

AB - This study aims to develop a reliable ground motion model (GMM) for CAV by using ground motion (GM) recordings from the PEER NGA-West2 database. A total of 17,684 GM recordings are chosen and randomly separated into the training, validation, and testing datasets. The DNN is advanced by incorporating the refined second-order (RSO) neuron. The effect of seismological and site-specific parameters on the predicted CAV is investigated. The comparative assessment of four existing models with the RSO-DNN model of this study highlights the superior prediction skill of the latter one since the RSO-DNN model is found to be associated with considerably less error.

KW - Cumulative absolute velocity

KW - Deep neural network

KW - Ground motion model

KW - PEER NGA-West2 database

KW - Standard deviation

U2 - 10.1080/13632469.2021.1985017

DO - 10.1080/13632469.2021.1985017

M3 - Journal article

SN - 1363-2469

VL - 26

SP - 8021

EP - 8040

JO - Journal of Earthquake Engineering

JF - Journal of Earthquake Engineering

IS - 15

ER -

Prediction of Cumulative Absolute Velocity Based on Refined Second-order Deep Neural Network

Abstract

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this