Banca de DEFESA: ADAM DREYTON FERREIRA DOS SANTOS

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
DISCENTE: ADAM DREYTON FERREIRA DOS SANTOS
DATA: 27/04/2017
HORA: 09:30
LOCAL: Auditório ITEC
TÍTULO:

OUTPUT-ONLY METHODS FOR DAMAGE IDENTIFICATION IN STRUCTURAL HEALTH MONITORING

PALAVRAS-CHAVES:

Structural health monitoring, Machine learning, Damage identification, Damage detection, Environmental conditions, Operational conditions.

PÁGINAS: 160
GRANDE ÁREA: Engenharias
ÁREA: Engenharia Elétrica
RESUMO:

In the structural health monitoring (SHM) field, damage identification based on the vibration response measurements from engineering structures has become a crucial research area due to its potential to be applied in real-world problems. Assuming that the vibration signals are often available and can be measured by employing different types of monitoring systems, when one applies appropriate data treatment, damage-sensitive features can be then extracted and used to assess early and progressive structural damage. However, real-world structures are subjected to regular changes in operational and environmental conditions (e.g. temperature, relative humidity, traffic loading and so on) which impose difficulties to identify structural damage as these changes influence different damage-sensitive features in a distinguish manner. Currently, the separation of changes in damage-sensitive features caused by damage from those caused by changing operational and environmental conditions remains as one of the major challenges to transit SHM technology from research to practice. In this thesis, to overcome this drawback, the SHM process is posed in the context of the statistical pattern recognition paradigm, where machine learning is the science of getting computers and algorithms to model the reality without knowing the physical laws of structures. This paradigm intends to pave the way for data-based models applicable to detect damage in structural systems of arbitrary complexity. The objective of this thesis is to adapt, develop, and apply several output-only methods based on machine learning and artificial intelligence for feature extraction and statistical modeling for feature classification capable of detect damage on structures under unmeasured operational and environmental influences. To test the performance of adapted and novel methods, as well as to compare they to state-of-the-art methods, the approaches are first applied on standard data sets measured from a laboratory three-story frame structure and then on response data from real-world structures – Z-24 and Tamar Bridges. The results demonstrated that the novel methods have better classification performance than the alternative ones in terms of false-positive and false-negative indications of damage, suggesting their applicability for real-world SHM solutions. If the proposed methods are compared to each other, the cluster-based ones, namely the global expectation-maximization approaches based on memetic algorithms, prove to be the best techniques to learn the normal structural condition without loss of information or sensitivity to the initial parameters and to detect damage (total errors equal to 4.4%)

MEMBROS DA BANCA:
Interno - 1698390 - ADAMO LIMA DE SANTANA
Externo ao Programa - 1809092 - CLAUDOMIRO DE SOUZA DE SALES JUNIOR
Interno - 3439716 - DIEGO LISBOA CARDOSO
Externo à Instituição - ELÓI JOÃO FARIA FIGUEIREDO
Presidente - 2153181 - JOAO CRISOSTOMO WEYL ALBUQUERQUE COSTA
Externo ao Programa - 1535051 - REGINA AUGUSTA CAMPOS SAMPAIO