No description available.

Dans cette thèse, nous développons une méthode d’estimation de l’effacement de consommation électrique d’un groupe de clients résidentiels. L’effacement, correspondant à une réduction de la puissance électrique sur une certaine durée, est désormais valorisé sur les marchés électriques et contribue à équilibrer le système électrique. Pour le quantifier, il faut estimer qu’elle aurait été la puissance appelée, i.e. la baseline, en l’absence de l’effacement. Ce dernier s’obtient alors par différence de la baseline et de la puissance réalisée. Les méthodes d’estimation de la baseline reposent sur des profils de consommation, des modèles de régression et des méthodes fondées sur un groupe de contrôle. Ces dernières offrent les résultats les plus précis mais déployer un groupe de contrôle aléatoire pour un usage opérationnel n’est pas envisageable.On s’intéresse donc à sélectionner un groupe de contrôle non-expérimental selon deux approches : la première emploie les caractéristiques observables des clients contrôles et la seconde leurs courbes de charge individuelles. Cette dernière idée consiste à sélectionner ces individus tels que la distance entre leur courbe de charge moyenne et celle du groupe recevant les effacements soit minimale. A cette fin, nous proposons un algorithme de sélection et adaptons les méthodes de régression sous contrainte, ridge et Lasso. Ces nouvelles méthodes procurent les meilleurs résultats. Enfin, pour estimer l’effacement en ligne, nous mettons en place un outil innovant qui associe un système de gestion de flux de données à un logiciel statistique / In this thesis, we develop a method in order to estimate the residential electricity demand reduction. The demand reduction or the curtailment, aiming at reducing the energy use during a short period, is currently enhanced on electricity markets and contributes to balance the electric system. To quantify it, one has to estimate the consumption, i.e. the baseline, which would have been used in the absence of the demand reduction. The curtailment is then obtained by subtracting the metered load during the demand reduction event from the baseline. The baseline estimation methods rely on day or weather matching methods, regression models and control group approaches. These one give the more accurate results but deploying a randomized control group is not possible for an operational use.We are then interested in selecting a non-experimental control group according to two approaches: the first uses the observable characteristics of the control customers and the second their individual load curves. This last idea consists in selecting those individuals such that the distance between their average load curve and that of the demand reduction group is minimal. To this end, we develop a forward selection algorithm and apply the constrained regression methods, ridge and Lasso. These methods provide the best results. Finally, we set up an innovative process which links a data flow module with a statistical software and allows to estimate the demand reduction on-line.

Light fields have been populated as a new geometry representation of 3D scenes, which is composed of multiple views, offering large potentials to improve the depth perception in the scenes. The light fields can be captured by different camera sensors, in which different acquisitions give rise to different representations, mainly containing a line of camera views - 3D light field representation, a grid of camera views - 4D light field representation. When the captured position is uniformly distributed, the outputs are the structured light fields. This thesis focuses on depth estimation from the structured light fields. The light field representations (or setups) differ not only in terms of 3D and 4D, but also the density or baseline of camera views. Rather than the objective of reconstructing high quality depths from dense (narrow-baseline) light fields, we put efforts into a general objective, i.e. reconstructing depths from a wide range of light field setups. Hence a series of depth estimation methods from light fields, including traditional and deep learningbased methods, are presented in this thesis. Extra efforts are made for achieving the high performance on aspects of depth accuracy and computation efficiency. Specifically, 1) a robust traditional framework is put forward for estimating the depth in sparse (wide-baseline) light fields, where a combination of the cost calculation, the window-based filtering and the optimization are conducted; 2) the above-mentioned framework is extended with the extra new or alternative components to the 4D light fields. This new framework shows the ability of being independent of the number of views and/or baseline of 4D light fields when predicting the depth; 3) two new deep learning-based methods are proposed for the light fields with the narrow-baseline, where the features are learned from the Epipolar-Plane-Image and light field images. One of the methods is designed as a lightweight model for more practical goals; 4) due to the dataset deficiency, a large-scale and diverse synthetic wide-baseline dataset with labeled data are created. A new lightweight deep model is proposed for the 4D light fields with the wide-baseline. Besides, this model also works on the 4D light fields with the narrow baseline if trained on the narrow-baseline datasets. Evaluations are made on the public light field datasets. Experimental results show the proposed depth estimation methods from a wide range of light field setups are capable of achieving the high quality depths, and some even outperform state-of-the-art methods. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

While substantial literature exists regarding the neurocognitive sequelae of concussion and return to play determinations for student-athletes who have suffered a concussion, there is a paucity of research that has conceptualized the impact of concussion on common academic outcomes. This study examined these topics in an attempt to evaluate the impact of concussion on GPA and school attendance, the association of preinjury neurocognitive performance to changes in academic achievement following a concussion, as well as the relationship between baseline neurocognitive performance and academic outcomes. The change in monthly assignment GPA and attendance were analyzed for three comparison groups (concussion, other sport-related injuries, control) regarding their differences pre and post injury. Second, whether a student-athlete's performance on a computerized baseline neuropsychological assessment (ImPACT) moderated the change in monthly assignment GPA for a group of concussed student-athletes identified as either short recovery or long recovery was investigated. Lastly, the association between baseline ImPACT scores and cumulative GPA/standardized achievement measures was examined for the entire group of student- athletes. Results of this study indicated no systematic differences between comparison groups (concussion, control, and injury) with the change in pre and post injury monthly assignment GPA and daily attendance rates, although academic declines were evident among all student-athletes. Further analysis revealed that more athletes who experienced concussions evidenced a significant drop in GPA (> .5) than would be expected by chance. Contrary to expectations, baseline ImPACT assessment scores did not moderate the degree of academic decline for concussed student-athletes within the short or long recovery groups. Finally, baseline ImPACT composite scores were significantly related to general academic achievement outcomes. Specifically, the ImPACT Visual Memory, Visual Motor and Reaction Time Composite scores significantly predicted GPA. Standardized academic achievement scores as measured by the Arizona Instrument to Measure Standards (math, reading, writing and science) were all significantly predicted by the baseline ImPACT Visual Motor Composite score. This study highlights the risk factors that may lead to diminished academic performance for student-athletes and the pre-injury neurocognitive variables measured by ImPACT that predict academic performance for student-athletes.

Demand participation is a basic ingredient of the next generation of power exchanges in electricity markets. A key challenge in implementing demand response stems from establishing reliable market frameworks so that purchasers can estimate the demand correctly, buy as economically as possible and have the means of hedging the risk of lack of supply. System operators also need ways of estimating responsive load behaviour to reliably operate the grid. In this context, two aspects of demand response are addressed in this study: scheduling and baseline estimation. The thesis presents a market clearing algorithm including demand side reserves in a two-stage stochastic optimization framework to account for wind power production uncertainty. The results confirm that enabling the load to provide reserve can potentially benefit consumers by reducing electricity price, while facilitating a higher share of renewable energy sources in the power system. Two novel methods, Bayesian Linear regression and Kernel adaptive filtering, are proposed for baseline load forecasting in the second part of the study. The former method provides an integrated solution for prediction with full accounting for uncertainty while the latter provides an online sequential learning algorithm that is useful for short term forecasting. / Graduate / 0544 / nimahtehrani@gmail.com

Thesis (M.Sc. Agric. (Horticulture)) --University of Limpopo, 2008 / The mineral nutrition of indigenous crop species is not well documented like other known crop species, thus making it difficult for one to know how to plant and maintain the crops. Mmupudu (Mimusops zeyheri), which, happens to be a wild crop, is one of the indigenous trees of interest to the Discipline of Plant Production, University of Limpopo. The current study gives baseline information that will be important in various environmental physiology studies of this plant. Physiological studies will be necessary to assess the importance of “limiting” mineral nutrients in the accumulation of certain mineral nutrients in Mmupudu in relation to its productivity. The experiment was arranged as a 2 x 3 factorial in RCBD, with the first and second factors being time of sampling and location, respectively. The three locations where data were collected were Chuenespoort, Bochum and Sekgosese. In each location, the experiment was replicated 10 times. Data were analyzed using ANOVA and means were separated using the least significant difference test. The two-factor interaction was nonsignificant (P ≥ 0.10) for both pH and electrical conductivity. Soil pH was not affected by time in all three locations suggesting that abscised flowers and fruitlets have no effect on pH. Leaf K experienced an increase of 65% at Chuenespoort and a decrease of soil K after fruiting by 44%. Leaf and soil P decreased after fruiting in all locations as was the case with Cu. Chuenespoort and Sekgosese experienced a decrease in leaf Mn after fruiting while soil Mn decreased in all whereas leaf Mg decreased in all locations. / the National Research Foundation and the Department of Water Affairs and Forestry

The medial olivocochlear reflex (MOCR) is a brainstem-mediated reflex that reduces cochlear amplifier gain when elicited by sound. The MOCR may provide benefits such as protection from acoustic trauma and improved hearing in background noise. Measurement of MOCR effects may also have clinical applications. MOCR effects can be measured using transient-evoked otoacoustic emissions (TEOAEs), as amplitudes of TEOAEs are typically reduced during MOCR activation. The primary purpose of the current study was to quantify the repeatability of MOCR effects on TEOAEs because high repeatability in a healthy population is a necessary (but not sufficient) component of a clinically-useful test. A secondary purpose was to assess the relationship between MOCR strength and speech perception in noise. Twenty-one normal hearing subjects ages 18-30 participated. TEOAEs were elicited using 35 dB sensation level (SL) clicks. The MOCR was elicited using contralateral acoustic stimulation (CAS) consisting of 35 dB SL broadband noise. Sixteen measurements were made across a 5-week period (4 visits × 4 measurements per visit). TEOAEs were bandpass filtered in 1/6-octaves from 1-2 kHz. An individualized time-frequency analysis was used to select the largest TEOAE envelope peak within a restricted time analysis window. Responses were characterized as the complex ratio of TEOAEs obtained with versus without CAS. The statistical significance of effects was assessed. Results revealed generally high levels of stability across time, as assessed by the interquartile ranges of all results and as assessed by Cronbach's alpha. Four MOCR measurements appeared to be adequate to obtain a reliable baseline measurement. Individualized time-frequency analyses were also important for obtaining reliable measurements. However, several subjects showed stable baseline measurements but unusual patterns of variability at subsequent sessions. These changes did not appear to be the result of changes in auditory status, methodological issues, or equipment issues. No significant relationship was found between MOCR strength and speech perception in noise. Results suggest that MOCR measurements are stable in most subjects when using careful measurement and analysis methods, but that further work is needed to better characterize changes in MOCR and to validate the current methodology in a larger number of subjects.

The glass sponge reefs of western Canada are modern analogues to ancient reefs and are unique habitats requiring conservation. However, the patterns and processes of the glass sponges have not been empirically studied. Here, I characterized the biology of the glass sponges in their reefs. I examined the community structure of the sponges at 3 reefs in the Strait of Georgia (SOG), their role in silica cycling, and the stable isotopes (13C and 15N) of the reef forming sponge Aphrocallistes vastus. Sponges are spatially structured in patches which localize the abundance of other animals. Long term dissolution of spicules is negligible and thus a reef can be considered a silica sink. Lastly, isotope compositions can differentiate populations of A. vastus and depleted carbon signatures at 2 reefs suggest a terrestrial component in their diet. My work represents the biological baseline of 3 glass sponge reefs in the SOG. / Ecology

Precise positioning is crucial for the success of navigation of underwater vehicles. At present, different instruments and methods are available for underwater positioning but few of them are reliable for three-dimensional position sensing of underwater vehicles. Long baseline (LBL) positioning is the standard method for three-dimensional underwater navigation. However, the accuracy of LBL positioning suffers from its own drawback of relatively low update rates. To improve the accuracy in positioning an underwater vehicle, integration of additional sensing measurements in a LBL navigation system is necessary. In this study, numerical simulation and experiment are conducted to investigate the effect of interrogate rate on the accuracy of LBL positioning. Numerical and experimental results show that the longer the interrogate rate, the greater the LBL positioning error. In addition, no reply from a transponder to transceiver interrogation is another major error source in LBL positioning. The experimental result also shows that the accuracy of LBL positioning can be significantly improved by the integration of velocity sensing. Therefore, based on Kalman filter, this study integrates a LBL system with vehicle dynamic model to improve the accuracy of positioning an underwater vehicle. For conducting the positioning experiments, a remotely operated vehicle (ROV) with dedicated Graphic User Interface (GUI) is designed, constructed, and tested. To have a precise motion simulation of ROV, a nonlinear dynamic model of ROV with six degrees of freedom (DOF) is used and its hydrodynamic parameters are identified. Finally, the positioning experiment is run by maneuvering the ROV to move along an ¡§S¡¨ trajectory, and Kalman filter is adopted to propagate the error covariance, to update the measurement errors, and to correct the state equation when the measurements of range, depth, and thruster command are available. The experimental result demonstrates the effectiveness of the integrated LBL system with the ROV dynamic model on the improvement of accuracy of positioning an underwater vehicle.

This dissertation develops a "neighborhood" based neural network model utilizing wavelet analysis and Self-organizing Map (SOM) to predict building baseline energy use. Wavelet analysis was used for feature extraction of the daily weather profiles. The resulting few significant wavelet coefficients represent not only average but also variation of the weather components. A SOM is used for clustering and projecting high-dimensional data into usually a one or two dimensional map to reveal the data structure which is not clear by visual inspection. In this study, neighborhoods that contain days with similar meteorological conditions are classified by a SOM using significant wavelet coefficients; a baseline model is then developed for each neighborhood. In each neighborhood, modeling is more robust without unnecessary compromises that occur in global predictor regression models. This method was applied to the Energy Predictor Shootout II dataset and compared with the winning entries for hourly energy use predictions. A comparison between the "neighborhood" based linear regression model and the change-point model for daily energy use prediction was also performed. We also studied the application of the non-parametric nearest neighborhood points approach in determining the uncertainty of energy use prediction. The uncertainty from "local" system behavior rather than from global statistical indices such as root mean square error and other measures is shown to be more realistic and credible than the statistical approaches currently used. In general, a baseline model developed by local system behavior is more reliable than a global baseline model. The "neighborhood" based neural network model was found to predict building baseline energy use more accurately and achieve more reliable estimation of energy savings as well as the associated uncertainties in energy savings from building retrofits.