Made available in DSpace on 2014-10-09T12:43:55Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T14:08:37Z (GMT). No. of bitstreams: 1 06786.pdf: 7146263 bytes, checksum: 0ad9a0e558c3dc7c1847019c4d7753a1 (MD5) / Dissertacao [Mestrado] / IPEN/D / Instituto de Pesquisas Energeticas e Nucleares - IPEN/CNEN-SP

Observa-se nos tempos atuais um crescente interesse e demanda por novas tecnologias de sensoriamento e interação. A monitoração, com o objetivo de reconhecimento de movimentos humanos, permite oferecer serviços personalizados em diferentes áreas, dentre elas a área de cuidados médicos. Essa monitoração pode ser realizada por meio de diferentes técnicas como o uso de câmeras de vídeo, instrumentação do ambiente onde o indivíduo habita, ou pelo uso de dispositivos pessoais acoplados ao corpo. Os dispositivos acoplados ao corpo apresentam vantagens como baixo custo, uso confortável, além de muitas vezes serem despercebidos pelo usuário, diminuindo a sensação de invasão de privacidade durante a monitoração. Além disso, o dispositivo sensor pode ser facilmente acoplado ao corpo pelo próprio usuário, tornando o seu uso efetivo. Deste modo, este trabalho apresenta o desenvolvimento de um sistema que emprega técnicas de inteligência computacional e um acelerômetro facilmente acoplado ao punho do usuário para efetuar, de maneira confortável e não invasiva, o reconhecimento de movimentos básicos da rotina de uma pessoa. Aplicando máquinas de vetores de suporte para classificar os sinais e a razão discriminante de Fisher para efetuar a seleção das características mais significativas, o sistema apresentou uma taxa de sucesso em torno de 93% no reconhecimento de movimentos básicos efetuados por indivíduos monitorados. O sistema apresenta potencialidade para ser integrado a um hardware embarcado de baixo custo, responsável pelo gerenciamento da aquisição dos dados e pelo encaminhamento das informações a um sistema de monitoramento ou armazenamento. As informações providas por este sistema podem ser destinadas à promoção da saúde e bem estar do indivíduo, bem como utilizadas em diagnósticos ou monitoramento remoto de pacientes em um ambiente de vida assistida. / Nowadays it is observed a growing interest and demand for new sensing technologies and interaction. Monitoring with the objective of recognizing human movements, allows us to offer personalized services in different areas, among them healthcare. This monitoring can be performed through the use of different techniques such as the use of video cameras, living environment instrumentation, or the use of personal devices attached to the body, also known as wearable devices. These wearable devices have some advantages such as low cost, comfortable to use, and are often unnoticed by the user, reducing the feeling of privacy invasion during the monitoring. In addition, the sensing device can be easily attached to the body by the user itself, making its use effective. Thus, this work presents the development of a system that uses computational intelligence techniques and an accelerometer which is easily attached to the users wrist to perform, in a comfortable and non-invasive manner, the recognition of basic movements of a persons routine. By applying support vector machines to classify the signals and Fishers discriminant ratio to select the most significant features, the system has shown a success rate of 93% in the recognition of basic movements performed by monitored individuals. The system has the potential to be integrated into a low-cost embedded hardware, which is responsible for managing the data acquisition and routing the movement data to a remote monitoring system or storage. The information provided by the system can be designed to promote the health and wellness of the individual, as well used in diagnostics or remote patient monitoring in an ambient assisted living (AAL).

La fragilité augmente dans le monde à cause de l’augmentation de la population âgée faisant passer l’être humain de l’autonomie à la dépendance. La détection précoce de la fragilité et le repérage des personnes à risque durant la phase de réversibilité de la fragilité permet d’engager rapidement des actions correctives. Le phénotype de Fried est l’outil le plus répandu pour l’identification de la personne fragile suivant cinq paramètres (vitesse de la marche, activité physique, perte de poids, fatigue et force de préhension palmaire). Un autre moyen de détection de la fragilité est d’évaluer la diminution de la complexité dans les signaux physiologiques. Cette diminution est traduite par la diminution de la corrélation à long terme dans les signaux physiologiques. Cependant, aucune approche ne permet d’alimenter ces indicateurs au domicile sans intervention d’un professionnel de santé. L’objectif de cette thèse est de contribuer à la définition d’un dispositif de mesure technologique, simple à utiliser et totalement intuitif, visant à alimenter deux indicateurs de Fried dans un environnement non contrôlé et détecter le signal de la marche sur une longue durée permettant la détection du changement dans la complexité du signal : le smartphone équipé d’un accéléromètre triaxial permet de mesurer l’activité physique et l’intervalle entre stride et stride qui est un outil important pour calculer la longueur de chaque pas, la vitesse de la marche et la variabilité dans le signal de la marche. Une stride est définit comme le temps entre le premier contact du talon avec le sol et le prochain contact de ce même talon / The frailty around the world is increasing because of the increase in the elderly population passing the human being from autonomy to dependence. Early detection of frailty and identification of individuals at risk during the reversibility phase of frailty provides quick initiations of corrective actions. The Fried phenotype is the most common tool for the identification of the fragile person using these five parameters (walking speed, physical activity, weight loss, fatigue and palmar grip strength). Another mean of the detection of frailty occurs in the reduction of the complexity in the physiological signals. The reduction in complexity is reflected in the decrement of the long-term correlation in temporal physiological signals. But for now, neither approach provides to supply to these indicators at home without the intervention of a health professional. The objective of this thesis is to contribute a technological measuring device, simple to use and totally intuitive to supply a few Fried indicators in an uncontrolled environment and detect the signal of walking for a long time allowing the detection of any change the complexity of the signal: the smartphone equipped with a tri axial accelerometer is used to measure the physical activity and the interval between stride and stride which is an important tool to calculate the length of each step, the walking speed and the variability in the signal of the step. A stride is defined as the time between the first contact of the heel with the ground and the next contact of the same heel

M.Ing. / A two-axis gimbal system in the form of a pitch-roll gimbal and a motion simulating platform were developed to extend the capability of an existing ground-to-air prototype gun-turret. The objective was to stabilize the line of sight (LOS) of the gimbal system despite disturbances introduced by the motion simulating platform in real time. The main sensor used for the stabilization is a two-axis accelerometer which was mounted directly on the inner gimbal (roll gimbal) to form a direct-mass stabilization architecture. The stabilization control algorithm was designed and executed in the Labview® environment on a PC, and the accelerometer data is used to drive the two DC motors used as the actuators of this control system. The design of the motion simulating platform was based on a simplified Stewart-Platform and uses pneumatic cylinders as actuating limbs. All sensors and actuators in the motion simulating platform and the gimbal system are integrated with the National Instrument’s CompactRio® and Labview®. The result was a simple stabilization controller capable of achieving basic stabilization of the LOS. However, the hardware and software of this project are capable of more complex control algorithms and that forms the bulk of the suggestions for further studies.

Gesture recognition plays an important role in human computer Interaction for intelligent computing. Major applications like Gaming, Robotics and Automated Homes uses gesture recognition techniques which diminishes the usage of mechanical devices. The main goal of my thesis is to interpret SWAT team gestures using different types of sensors. Accelerometer and flex sensors were explored extensively to build a prototype for soldiers to communicate in the absence of line of sight. Arm movements were recognized by flex sensors and motion gestures by Accelerometers. Accelerometers are used to measure acceleration in respect to movement of the sensor in 3D. Flex sensors changes its resistance based on the amount of bend in the sensor. SVM is the classification algorithm used for classification of the samples. LIBSVM (Library for Support Vector Machines) is integrated software for support vector classification, regression and distribution estimation which supports multi class classification. Sensors data is connected to the WI micro dig to digitize the signal and to transmit it wirelessly to the computing device. Feature extraction and Signal windowing were the two major factors which contribute for the accuracy of the system. Mean Average value and Standard Deviation are the two features considered for accelerometer sensor data classification and Standard deviation is used for the flex sensor analysis for optimum results. Filtering of the signal is done by identifying the different states of signals which are continuously sampled.

The objectives of this dissertation include 1) to review accelerometry and Global Navigation Satellite System (GNSS) derived measures used to monitor training load, 2) to investigate the validity and reliability of accelerometers (ACCs) to identify stepping events and quantify training load, 3) to assess the relationship between accelerometry and Global Navigation Satellite Systems (GNSS) derived measures in quantifying training load. In Study I, acceleration data was collected via two tri-axial ACC (Device A and Device B) sampling at 100Hz mounted closely together at the xiphoid process level. Each participant (n=30) performed two trials of straight-line walking, running, and sprinting on a 20m course. Device A was used to assess ACC validity to identify steps and the test-retest reliability of the instrument to quantify the external load. Device A and Device B were used to assess inter-device reliability. The reliability of accelerometry derived metrics Impulse Load (IL) and Magnitude g (MAG) were assessed. In Study II, known distance (DIST) was predicted via acceleration data collected by a tri-axial ACC sampling at 100Hz mounted at the xiphoid process level, simultaneously positional data collected using a triple GNSS unit sampling at 10Hz placed between scapulae. Each participant (n=30) walked different DIST around various movement constraints (small and large circles). Thirty distances were completed around each circle and ranged from 12.57–376.99m. In Study I, the instrument demonstrated a positive predictive value (PPV) of 96.98-99.41% and an agreement of 93.08-96.29% for step detection during all conditions. Good test-retest reliability was found with a coefficient of variation (CV) < 5% for IL and MAG during all locomotor conditions. Good inter-device reliability was also found for all locomotor conditions (IL and MAG CV < 5%). These results indicated that tri-axial ACCs are a valid and reliable tool used to identify steps and quantify external load when movement is completed at a range of speeds. In Study 2, all linear regression models performed well for both movement constraints (R2=0.922-0.999; RMSE=0.047-0.242, p

An investigation into the spatial ecology and effects of catch-and-release angling on the physiology and behavior of striped bass was conducted. Fine-scale behavior was assessed by tagging fish with acoustic transmitters equipped with pressure and tri-axial accelerometer sensors and tracking them within a fixed array (n=34 receivers) in a Massachusetts estuary. Activity space changed significantly over the course of the season and increased with water temperature. Striped bass most frequently exhibited low levels of locomotory activity representing 67% of total activity measurements, with occasional high activity and burst swimming, often within the upper 3 m of the water column. Depth distribution of striped bass remained shallower when temperatures peaked at over 21 oC. Diel vertical migration was present with shallowest depths observed during the day and greatest depths during high tide. To investigate catch-and-release consequences, 102 striped bass were angled and blood sampled between July and November 2011. A subsample of 35 striped bass (July n=11, August n=11, September n=13) were implanted with tri-axial acoustic accelerometers to assess relative behavior and survival post-release. Results from principle component analyses produced five factors describing 72.7% of the variance for blood physiology parameters, total length, and water temperature. Subsequently, only eigenvalues from PC1, with high loading for blood lactate, plasma sodium and chloride, and total length, were significantly correlated with fight time. Eight individual fish were detected within 12 hours of release and exhibited their greatest mean daily activity space estimate within that time (1.5 km2 ± 0.6, 50%; 5.6 km2 ± 2.2, 95%). Depth ranged from 0-6.15 m (1.89±1.3 m) and acceleration ranged from 0.095-3.51 ms-2 (0.95±0.33). In summary, no observed mortality suggests that fish were able to recover from the physical and physiological impacts of angling. This thesis has increased the understanding of striped bass ecology and will help promote future conservation and management initiatives for striped bass and facilitate additional research.

<p>Motivated by the future prospects of the Windchime project, we show that even a small prototype array of 7 commercial accelerometers can be used to calculate dark matter limits for the well-known B − L coupled dark photon. As a member of the ultralight sector, the dark photon would be observed in high occupancy as a persistent plane wave characterized by de Broglie wavelength and coupling to the standard model via a hypothesized baryon minus lepton quantum number, g_B−L. Such an interaction can be probed by measuring the differential force or acceleration between two bodies of differing B −L charge-to-mass ratios. This is accomplished for a 7 sensor array of MEMS accelerometers by rigidly coupling all the sensors to a material of known B − L charge. Using a log-likelihood ratio test and Fourier transformed data from the prototype array, we are able to set a limit on g_B−L ∼ 10^−11 within a mass range of 10^{−13}−10^{−12}eV . Setting these noncompetitive limits with real data serves as a proof-of-principle demonstration of the limit-setting procedure used in Windchime future projections for B − L coupled ultralight dark matter. Additionally, this basic setup could be used for future studies on the properties of a detector array. </p>

: Previous research has explored associations between accelerometry and Global Navigation Satellite System (GNSS) derived loads. However, to our knowledge, no study has investigated the relationship between these measures and a known distance. Thus, the current study aimed to assess and compare the ability of four accelerometry based metrics and GNSS to predict known distance completed using different movement constraints. A correlational design study was used to evaluate the association between the dependent and independent variables. A total of 30 physically active college students participated. Participants were asked to walk two different known distances (DIST) around a 2 m diameter circle (small circle) and a different distance around an 8 m diameter circle (large circle). Each distance completed around the small circle by one participant was completed around the large circle by a different participant. The same 30 distances were completed around each circle and ranged from 12.57 to 376.99 m. Acceleration data was collected via a tri-axial accelerometer sampling at 100 Hz. Accelerometry derived measures included the sum of the absolute values of acceleration (SUM), the square root of the sum of squared accelerations (MAG), Player Load (PL), and Impulse Load (IL). Distance (GNSSD) was measured from positional data collected using a triple GNSS unit sampling at 10 Hz. Separate simple linear regression models were created to assess the ability of each independent variable to predict DIST. The results indicate that all regression models performed well (R = 0.960-0.999, R = 0.922-0.999; RMSE = 0.047-0.242, < 0.001), while GNSSD (small circle, R = 0.999, R = 0.997, RMSE = 0.047 < 0.001; large circle, R = 0.999, R = 0.999, RMSE = 0.027, < 0.001) and the accelerometry derived metric MAG (small circle, R = 0.992, R = 0.983, RMSE = 0.112, < 0.001; large circle, R = 0.997, R = 0.995, RMSE = 0.064, < 0.001) performed best among all models. This research illustrates that both GNSS and accelerometry may be used to indicate total distance completed while walking.

The low size and budget of typical nanosatellite missions limit the available sensors for attitude estimation. Relatively high noise MEMS gyroscopes often must be employed when accurate knowledge of the spacecraft’s angular velocity is necessary for attitude determination and control. This thesis derived and tested in simulation the “Virtual Gyroscope” algorithm, which replaced a standard gyroscope with an array of spatially distributed accelerometers for a 1U CubeSat mission. A MEMS accelerometer model was developed and validated using Root Allan Variance, and the Virtual Gyroscope was tested both in the open loop configuration and as a replacement for a gyroscope in a Multiplicative Extended Kalman Filter. It was found that the quality of the Virtual Gyroscope’s rate measurement improved with a larger and higher quality array, but the error in the estimate was very large. The low signal-to-noise ratio and the unknown bias in the accelerometers caused the angular velocity estimate from the accelerometer array to be too poor for use in the propagation step of the Kalman filter. The Kalman filter performed better with attitude measurements alone than with the Virtual Gyroscope, even when the attitude were delivered at a low rate with added noise. Overall, the current Virtual Gyroscope algorithm that is presented in this thesis is not suitable to replace a MEMS gyroscope in a nanosatellite mission, although there is room for future improvements using bias prediction for the individual accelerometers in the array.