In the last decades, long-term structural health monitoring of civil structures has been mainly performed using two approaches: model based and data based. The former approach tries to identify damage by relating prediction numerical (e.g., finite-element) models structure. latter is driven, where measured from a given state condition are compared baseline or reference condition. A challenge in both approaches make distinction between changes response caused environmental operational...

In most real-world monitoring scenarios, the lack of measurements from damaged conditions requires application unsupervised approaches, mainly ones based on modal features estimated raw vibration data through traditional system identification methods. Although numerous successful applications using parameters have been reported, they demonstrated to be insufficient estimate a robust set damage-sensitive features. Inspired by idea compressed sensing and deep learning, an intelligent two-level...

During the service life of bridges, bridge management systems (BMSs) seek to handle all performed assessment activities by controlling regular inspections, evaluations, and maintenance these structures. However, BMSs still rely heavily on qualitative visual which compromise structural evaluation and, consequently, decisions as well avoidance collapses. The health monitoring appears a natural field aid management, providing more reliable quantitative information. Herein, machine learning...

The structural health monitoring relies on the continuous observation of a dynamic system over time to identify its actual condition, detect abnormal behaviors, and predict future states. regular changes in environmental factors have been reported as one main challenges for application systems. These influences responses are general nonlinear, affecting damage-sensitive features most varied forms. usual process remove these normal is referred data normalization. In that regard, principal...

During the service life of engineering structures, structural management systems attempt to manage all information derived from regular inspections, evaluations and maintenance activities. However, still rely deeply on qualitative visual which may impact evaluation and, consequently, decisions as well avoidance collapses. Meanwhile, health monitoring arises an effective discipline aid management, providing more reliable quantitative information; herein, machine learning algorithms have been...

This paper proposes a novel unsupervised damage detection approach based on memetic algorithm that establishes the normal or undamaged condition of structural system as data clusters through global expectation-maximization technique, using only damage-sensitive features extracted from output-only vibration measurements. The health state is then discriminated by considering Mahalanobis squared distance between learned and new observation. proposed compared with state-of-the-art ones taking...

The goal of this work is to detect structural damage using vibration-based identification approaches even when the damage-sensitive features are camouflaged by presence operational and environmental conditions. For feature classification purposes, four machine learning algorithms applied based on principal component analysis (PCA), nonlinear PCA, kernel PCA greedy PCA. Time-series data from an array accelerometers under several state conditions were obtained a well-known base-excited...

Improved and more continuous condition assessment of bridges has been demanded by our society to better face challenges presented aging civil infrastructure. In the last two decades, bridge techniques have developed in order improve maintenance a systematic way. The Structural Health Monitoring (SHM) given ability provide information, real time, about performance structural system. However, reliability that information depends highly on quality data analysis, as operational environmental...

O setor da Segurança Pública tem adotado as tendências de aplicação ciência e mineração dados, impulsionado pelo volume dados gerados diariamente pela automatização aprimoramento dos processos internos. Este estudo propõe a modelos linguagem baseados em BERT para classificar crimes relatos boletins ocorrências Marabá, Pará. Os resultados mostraram que RoBERTa alcançaram melhores performances, com acurácia entre 89% 90% dez classes relacionadas maior ocorrência. Tal automação classificação...

The process of verbally reporting or manually retyping numeric data generated at dual-energy x-ray absorptiometry (DXA) involves numerous pitfalls. With use a macro scripting editor, customized was created to automate the transfer by DXA scanner into structured voice recognition dictation system without requiring radiologists type in medical record number accession identify study. A preliminary report is with software for unit and template that includes qualitative assessments osteoporosis...

Convolutional neural networks (CNNs) are one of the most promising techniques from computer vision that can generate substantial gains in varied classification problems. In sense, this paper aims to perform image processing on XDB plant disease database for enhancing 20 different diseases 10 distinct species. This encompassed two pre-processing activities (image selection and resizing) application modified VGG architectures, VGG16 VGG19, along with pre-trained weights ImageNet database. A...

The rapid growth of the Internet and technological advances are forcing mobile operators to increasingly invest in network infrastructures. C-RAN SDN regarded as enabling technologies that can overcome limitations faced by operators, reducing costs, increasing scalability, paving way for next generation 5G cellular networks. In this paper, an architectural solution based on computational intelligence is proposed C-RAN, which adjust BBU-RRH mapping through load balancing rules predicting...

Pneumonia is one of the most common medical problems in clinical practice and leading fatal infectious disease worldwide. According to World Health Organization, pneumonia kills about 2 million children under age 5 constantly estimated be cause infant mortality, killing more than AIDS, malaria, measles combined. A key element diagnosis radiographic data, as chest x-rays are routinely obtained a standard care can aid differentiate types pneumonia. However, rapid radiological interpretation...

This paper aims to compare the convolutional neural networks(CNNs): ResNet50, InceptionV3, and InceptionResNetV2 tested withand without pre-trained weights on ImageNet database in orderto solve scene recognition problem. The results showed that thepre-trained ResNet50 achieved best performance with an averageaccuracy of 99.82% training 85.53% test, while theworst result was attributed pre-training,with 88.76% 71.66% average accuracy testing,respectively. main contribution this work is direct...

Optical sensors have found application in many fields, such as Civil Engineering, Aeronautics, Energy and Oil & Gas Industries. Monitoring solutions based on this technology proven particularly cost effective can be applied to large scale structures where hundreds of must deployed for long term measurements different mechanical physical parameters. Sensors Fiber Bragg Gratings (FBGs) are the most common solution used Structural Health (SHM) performed by instruments known optical...