Premium
Automated damage location for building structures using the hysteretic model and frequency domain neural networks
Author(s) -
MoralesValdez Jesús,
LopezPacheco Mario,
Yu Wen
Publication year - 2020
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.2584
Subject(s) - convolutional neural network , frequency domain , structural health monitoring , principal component analysis , displacement (psychology) , artificial neural network , computer science , time domain , structural engineering , domain (mathematical analysis) , building model , engineering , artificial intelligence , simulation , computer vision , mathematics , psychology , psychotherapist , mathematical analysis
Summary This paper presents a novel and accurate model‐reference health monitoring system for the location of damage to building structures using the dissipated energy approach, frequency domain convolutional neural networks (CNNs), and principal component analysis (PCA). Due to the fact that the earthquake introduces several stress cycles in different directions in the structure, load–strain curves can be used as an indicator of damage. The CNN in the frequency domain (CNNFI) is used to estimate the hysteretic displacement of the reference of the Bouc–Wen model. Automated damage locations are resolved with the CNN classification models (CNNFC). The comparison study for damage location is presented by using classical neural networks. The results of the damage location of a two‐story building prototype confirmed that the proposed method is promising for real applications.