No description available.

No description available.

No description available.

Recent years have seen increasing research interest in structural health monitoring (SHM). Among the many advances in SHM research, “smart” wireless sensors capable of embedded computing and wireless communication have been highly attractive. Wireless communication in SHM systems was originally proposed to significantly reduce the monetary and time cost for installing lengthy cables in an SHM system. Besides wireless sensing, the next revolution in sensor networks has been predicted to be mobile sensor networks that implant mobility into traditional wireless sensor networks. This research explores decentralized structural model updating and damage detection using mobile and wireless sensors. In the first stage of this research, mobile sensing nodes (MSNs) are developed for SHM purposes. The MSNs can maneuver upon structures built with ferromagnetic/steel materials, conduct measurement, and communicate with pears or remote servers wirelessly. The performance of the MSNs is validated through laboratory and field experiments. To further investigate the mobile sensing strategy, a decentralized structural damage detection procedure is proposed herein for the MSNs using transmissibility functions. Laboratory experiments are conducted on a steel portal frame where various structure damage scenarios are emulated. Besides experiments with MSNs, this study also investigates the nature of transmissibility functions for damage detection in an analytical manner based on a general multi-DOF spring-mass-damper system. Finally, this research also explores substructure model updating through minimization of modal dynamic residuals, which can best benefit from dense mobile or wireless sensor data concentrated in one area. Craig-Bampton transform is adopted to condense the structural model, and minimization of the modal dynamic residuals is determined as the optimization objective. An iterative linearization procedure is adopted for efficiently solving the optimization problem. The presented substructure updating method is validated through a few numerical examples. For comparison, a conventional approach minimizing modal property differences is also applied, and shows worse updating accuracy than the proposed approach. The performance of the proposed substructure model updating approach is further investigated on the effects of substructure location and size.

In partnership with iMonSys - an emerging company in the passive care field - a new system, 'Verity', is being developed to fulfil the role of a passive behaviour monitoring and alert detection device, providing an unobtrusive level of care and assessing an individual's changing behaviour and health status whilst still allowing for independence of its elderly user. In this research, a Hidden Markov Model incorporating Fuzzy Logic-based sensor fusion is created for the behaviour detection within Verity, with a method of Fuzzy-Rule induction designed for the system's adaptation to a user during operation. A dimension reduction and classification scheme utilising Curvilinear Distance Analysis is further developed to deal with the recognition task presented by increasingly nonlinear and high dimension sensor readings, and anomaly detection methods situated within the Hidden Markov Model provide possible solutions to identification of health concerns arising from independent living. Real-time implementation is proposed through development of an Instance Based Learning approach in combination with a Bloom Filter, speeding up the classification operation and reducing the storage requirements for the considerable amount of observation data obtained during operation. Finally, evaluation of all algorithms is completed using a simulation of the Verity system with which the behaviour monitoring task is to be achieved.

Made available in DSpace on 2016-04-04T18:31:39Z (GMT). No. of bitstreams: 1 LUIZ RODOLFO BARRETO DA SILVA.pdf: 1309393 bytes, checksum: c820374f746f8a21bb89acd6347d0907 (MD5) Previous issue date: 2014-02-07 / This work discusses the development of a method for estimating the distance of sensor nodes in Wireless Sensor Networks (WSN) in indoor environments, considering the need for adaptation of the method due to the presence of random factors that may influence the estimation of such distances. The estimation method presented is based on a signal strength indicator, RSSI (Received Signal Strength Indicator). / Este trabalho aborda a elabora??o de um m?todo para a estimativa de dist?ncia de n?s sensores em Redes de Sensores Sem Fio (RSSF) em ambientes internos fechados, considerando necessidade de adapta??o do m?todo devido ? presen?a de fatores aleat?rios que podem influenciar a estimativa dessas dist?ncias. O m?todo de estimativa apresentado ? baseado em um indicador de intensidade de sinal, o RSSI (Received Sinal Strength Indicator).

[pt] Algumas características básicas das redes de sensores sem fio (RSSF) dificultam as tarefas de criação e reconfiguração de aplicações. Nesse trabalho apresentamos um modelo de programação que pretende simplificar essas tarefas. O modelo se baseia no uso conjunto de funções parametrizáveis e de máquinas de estados finitos, e permite a implementação de diferentes tipos de aplicações para redes de sensores sem fio e a configuração remota dessas aplicações. Descrevemos alguns testes para avaliar o quanto esse modelo pode facilitar o desenvolvimento de novas aplicações, o quanto é fácil aplicar novas alterações sobre as aplicaçõesem execuçãos, e o impacto na quantidade de mensagens na rede por conta do uso da configuração remota. / [en] Some basic characteristics of wireless sensor networks (WSN) make application creation and reconfiguration dificult tasks. A programming model is presented to simplify these tasks. This model is based on a set of parametrized components and on a Finite State Machine, and allows the remote configuration of different applications over the same set of installed components. We describe some tests to evaluate its impact on the development process, and the ease of applying modifications to a running application. We also measure the additional impact of remote configuration on network activity.

A realtime wireless motion tracking system is developed. The system is capable of tracking the orientations of multiple wireless sensors, using a semi-distributed implementation to reduce network bandwidth and latency, to produce real-time animation of rigid body models, such as the human skeleton. The system has been demonstrated to be capable of full-body posture tracking of a human subject using fifteen devices communicating with a basestation over a single, low bandwidth, radio channel. The thesis covers the theory, design, and implementation of the tracking platform, the evaluation of the platform’s performance, and presents a summary of possible future applications.

The Kahne rumen sensor (bolus) is a device developed to measure temperature, pressure and pH in non-fistulated animals. This bolus allows real-time monitoring of the rumen environment, which could help preventing health problems such as rumen acidosis in cows. It is less invasive to use boluses compared to other technologies that measure the ruminal pH (e.g. rumenocentesis). Kahne boluses and transceivers are commercially available in the market. Several studies on the relationships between data recorded by the bolus and actual data recorded by independent devices were conducted. The bolus temperature and pressure were compared with actual temperature and pressure under controlled conditions. The pH drift was studied by comparing the difference between bolus and direct measurement over time. The capture of the data was calculated for each bolus in various experiments to examine the factors affecting the data capture rate of the boluses. Animal to animal variation was studied using boluses in a group of cows fed and managed under uniform conditions. An animal experiment involving fistulated cows eating two different diets was performed using boluses to monitor the changes of ruminal pH. There was no apparent interruption to normal animal behaviour as a result of using boluses. Regression relationships between bolus measurements and actual data for both temperature and pressure were developed and used for calibration of bolus data. The pH drift was a problem, as the regression relation between the pH difference and the time for one bolus from one experiment could not represent this bolus on other experiment. The data capture rate on the hourly basis ranged from 0 to 100%, but was usually between 30 to 70%. The data capture rate was affected by many factors and further studies to identify these factors are needed. A study of animal to animal variation suggests that in a comparison of 2 treatments, a minimum 3 cows per group would be required to detect the standard deviation of 0.11 for a pH difference of 5% of the mean (approximately 0.35 pH units). Seventeen cows per group would be required to detect the standard deviation of 0.33 for the same difference. The boluses effectively monitored the ruminal pH change in cows easting two different diets and the profile of change of pH was successfully analysed. Feeding 7.6 kg baleage twice a day cause pH to decrease at 0.009 pH units per minute during feeding, while offering a similar quantity of grass and hay once a day resulted in a decrease of 0.0009 pH units per minute during feeding. The beginning of pH increase was about 1 hour following feeding and continuous during resting and rumination. The level of pH increase did not differ significantly for two diets.. The Kahne devices appear to have advantages compared to other technologies for the measurement of parameters of the rumen environment on a real-time basis. Boluses are especially good at intensively monitoring the temperature, pressure, and pH in the rumen. The major limitations of this technology to be used are the data capture rate and the pH drift. By improving the limitations found in the experiment, the Kahne rumen sensor could become very useful for both scientific research and under commercial conditions for monitoring animal health.

Made available in DSpace on 2015-04-11T14:02:36Z (GMT). No. of bitstreams: 1 Felipe Lobo.pdf: 609088 bytes, checksum: ab4ee5abe3ca57cd1089d41d9089b070 (MD5) Previous issue date: 2012-03-22 / One of the main applications on Wireless Sensors Networks (WSNs) is the environmental monitoring, particularly in monitoring of the fauna in diverse places like forests, fields, etc. In these environments that are of difficult access, the maintenance of the sensors is impaired and energy restriction becomes a major factor in the life of the network. How in most such cases the data are not needed in real time, in this work is proposed and evaluated a new cross-layer algorithm, called PWAVE, turned to WSNs with periodic data collection, where data transmissions are made through a predetermined interval. While such transfers are not performed, the sensor still collecting information from the environment and turn off their radios to reduce energy consumption. Several other challenges appear before this way of working, for example, the synchronization of clocks, which are readily solved by the algorithm. The results show that the proposed solution is efficient and can save up to 51.2 % of energy with 99 % in the rate of data delivery when compared with similar solutions / Uma das principais aplicações existentes em Redes de Sensores Sem Fio (RSSFs) é o monitoramento ambiental, mais particularmente no acompanhamento da fauna existente em locais diversos como florestas, campos, etc. Nesses ambientes de difícil acesso, a manutenção dos sensores fica prejudicada e a restrição de energia torna-se um fator primordial para a vida útil da rede. Como na maioria desses casos os dados não são necessários em tempo real, neste trabalho é proposto e avaliado um novo algoritmo cross-layer, denominado PWAVE, voltado para RSSFs com coleta periódica, onde as transmissões dos dados são feitas através de intervalos predeterminados. Enquanto essas transmissões não são executadas, o sensor permanece coletando as informações do ambiente e seus rádios desligam-se, a fim de reduzir o consumo de energia. Diversos outros desafios aparecem diante deste modo de trabalho como, por exemplo, a sincronização dos relógios, que são prontamente solucionados pelo algoritmo proposto. Os resultados mostram que a solução proposta é eficiente e consegue economizar até 51.2% de energia com 99% na taxa de entrega de dados quando comparado com soluções similares