An improvement on feature extraction via time series modeling for structural health monitoring based on unsupervised learning methods

Entezami, Alireza; Shariatmadar, Hashem; Karamodin, Abbas

doi:10.24200/sci.2018.20641

Cited by 11 publications

(10 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As Figure 7 presents, a considerable part of the hot-important topics in the IT domain is allocated for the Health. This figure indicates that publications of this domain have been shifting from pure computational topics (such as Data Mining and Machine Learning) to application of computer science in the other domains (especially in the health domain) for example [75][76][77]. Moreover, In the last years, individuals working at the intersection of IT and medicine have been developed and computer applications to improve health care services have increased [16].…”

Section: E | Portion = |P|mentioning

confidence: 99%

Evolution of IT, Management and Industrial Engineering research: a topic model approach

et al. 2020

View full text Add to dashboard Cite

Information Technology (IT), Management and Industrial Engineering are correlated academic disciplines which their publications rose significantly over the last decades. The aim of this study is analyzing the research evolution, determining the important topics and areas and depiction the trend of interdisciplinary topics in these domains. To accomplish this, the text mining techniques are used and the combination of bibliographic analysis and topic modeling approach are applied on their publications in the WOS repository over the last 20 years. In the topic extraction process, a heuristic function was suggested to key extraction, and some new applicable criteria were defined to compare the topics. Moreover, a novel approach was proposed to determine the high-level category for each topic. The results determined the hot-important topics and incremented, decremented and fixed topics are identified. Subsequently, comparing the high-level research areas confirmed the strong scientific relationships between them. This study presents a deep knowledge about internal research evolution of domains and illustrates the effect of topics on each other over the past 20 years. Furthermore, the methodology of this study could be applied to determine the interdisciplinary topics and observe the research evolution of other academic domains.

show abstract

Section: E | Portion = |P|mentioning

confidence: 99%

Evolution of IT, Management and Industrial Engineering research: a topic model approach

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Due to recent advances in sensing and data acquisition systems, the strategies in the SHM realm have been shifted from model-driven techniques under the concept of finite element model updating [ 10 , 11 , 12 , 13 ] to data-driven or data-based methods based on statistical pattern recognition and machine learning [ 1 , 14 , 15 , 16 , 17 ]. In contrast to model-based techniques that require elaborate numerical models of real-life structures, data-driven methods are only based on raw measurements with no requirement for numerical modeling and model updating strategies.…”

Section: Introductionmentioning

confidence: 99%

Damage Detection in Largely Unobserved Structures under Varying Environmental Conditions: An AutoRegressive Spectrum and Multi-Level Machine Learning Methodology

Entezami

Mariani

Shariatmadar

2022

Sensors

Self Cite

View full text Add to dashboard Cite

Vibration-based damage detection in civil structures using data-driven methods requires sufficient vibration responses acquired with a sensor network. Due to technical and economic reasons, it is not always possible to deploy a large number of sensors. This limitation may lead to partial information being handled for damage detection purposes, under environmental variability. To address this challenge, this article proposes an innovative multi-level machine learning method by employing the autoregressive spectrum as the main damage-sensitive feature. The proposed method consists of three levels: (i) distance calculation by the log-spectral distance, to increase damage detectability and generate distance-based training and test samples; (ii) feature normalization by an improved factor analysis, to remove environmental variations; and (iii) decision-making for damage localization by means of the Jensen–Shannon divergence. The major contributions of this research are represented by the development of the aforementioned multi-level machine learning method, and by the proposal of the new factor analysis for feature normalization. Limited vibration datasets relevant to a truss structure and consisting of acceleration time histories induced by shaker excitation in a passive system, have been used to validate the proposed method and to compare it with alternate, state-of-the-art strategies.

show abstract

“…Feature extraction aims to extract meaningful information from the measured vibration data that should be correlated with damage, known as a damage-sensitive feature (DSF) [13]. Time series analysis [14][15][16], time-frequency signal analysis [17][18][19], principal component analysis (PCA) [20][21][22][23] are widely-used and effective methods for feature extraction. Feature analysis is a decision-making procedure that utilizes the DSFs of undamaged and damaged conditions extracted from vibration signals in order to analyze them for early damage detection and damage localization.…”

Section: Introductionmentioning

confidence: 99%

Identifying Damage Location under Statistical Pattern Recognition by New Feature Extraction and Feature Analysis Methods

Feizi,

Sharbatdar,

Mahjoub

et al. 2021

Scientia Iranica

View full text Add to dashboard Cite

The main objective of this article is to identify the location of damage by a new feature extraction technique and propose some efficient feature analysis tools as statistical distance measures. The proposed algorithm of feature extraction relies on a combination of the well-known Principal Component Analysis (PCA) and a convolution strategy. After extracting the features from raw vibration signals of undamaged and damaged conditions, those are applied to the proposed feature analysis approaches called coefficient of variation, Fisher criterion, Fano factor, and relative reliability index, all of which are formulated by using statistical moments of the features extracted from the PCA-convolution algorithm. To localize damage, the sensor location with the distance value exceeded from a threshold limit is identified as the damaged area. The main innovations of this research are to present a new hybrid technique of feature extraction suitable for SHM applications and four effective statistical measures for feature analysis and damage identification. The performance and reliability of the proposed methods are verified by a four-story shear-building model and a benchmark concrete beam. Results demonstrate that the approaches presented here can influentially identify the location of damage by using the features extracted from the proposed PCA-convolution algorithm.

show abstract

An improvement on feature extraction via time series modeling for structural health monitoring based on unsupervised learning methods

Cited by 11 publications

References 24 publications

Evolution of IT, Management and Industrial Engineering research: a topic model approach

Evolution of IT, Management and Industrial Engineering research: a topic model approach

Damage Detection in Largely Unobserved Structures under Varying Environmental Conditions: An AutoRegressive Spectrum and Multi-Level Machine Learning Methodology

Identifying Damage Location under Statistical Pattern Recognition by New Feature Extraction and Feature Analysis Methods

Contact Info

Product

Resources

About