Reliable, efficient and cost-effective energy supply is crucial for economic and social development. Mining and industrial sectors consumed close to 37% of the total energy produced in the world during 2013. The South African power network is strained by the rapid expansion of mining, industrial and public sectors. Generation, transmission and distribution of electrical energy are in progress, but supply will not meet demand in the near future. The South African electricity supplier needs capital for expansion. Electricity price increases have been significantly higher than increases in the gold price over the last few years. Mining companies are under pressure from government to improve their labour relations. They are obligated to spend money on local infrastructure development. Therefore, cost efficiency receives higher priority than ever before and requires an implementation strategy. Cooling systems on mines proved to be significant electricity consumers. These systems lack integrated management and efficient and optimised control. Electricity demand can be reduced through implementation of energy saving measures on these cooling systems. Energy saving measures reduce the operational costs of mining to ensure that mines stay globally competitive. The identification of long-term challenges for energy saving measures is crucial. Successful implementation of energy saving measures results in improved utilisation and performance of mine cooling systems. These measures must be maintained to ensure a constant positive impact on reduced electrical energy consumption. The electrical energy savings are dependent on external factors, such as ambient conditions. Improved implementation strategies of energy saving measures will prevent deterioration of utilisation and performance of the mine cooling systems. Monitoring and reporting of key performance indicators are crucial. Lack of integrated maintenance can lead to lost opportunities and the deterioration of equipment and machines. The improved implementation strategies in two separate case studies proved sustainable savings of 1.73 MW and 0.66 MW respectively. The electricity cost savings for Mine A and Mine B are R8.8 million and R2.9 million respectively. These savings have been sustained for periods of seventeen and seven months respectively, indicating the value of the study. / MIng (Mechanical Engineering), North-West University, Potchefstroom Campus, 2015

<p>Collections of large volumes of rich and complex data has become ubiquitous in recent years, posing new challenges in methodological and theoretical statistics alike. Today, statisticians are tasked with developing flexible methods capable of adapting to the degree of complexity and noise in increasingly rich data gathered across a variety of disciplines and settings. This has spurred the need for novel multivariate regression techniques that can efficiently capture a wide range of naturally occurring predictor-response relations, identify important predictors and their interactions and do so even when the number of predictors is large but the sample size remains limited. </p><p>Meanwhile, efficient model fitting tools must evolve quickly to keep pace with the rapidly growing dimension and complexity of data they are applied to. Aided by the tremendous success of modern computing, Bayesian methods have gained tremendous popularity in recent years. These methods provide a natural probabilistic characterization of uncertainty in the parameters and in predictions. In addition, they provide a practical way of encoding model structure that can lead to large gains in statistical estimation and more interpretable results. However, this flexibility is often hindered in applications to modern data which are increasingly high dimensional, both in the number of observations $n$ and the number of predictors $p$. Here, computational complexity and the curse of dimensionality typically render posterior computation inefficient. In particular, Markov chain Monte Carlo (MCMC) methods which remain the workhorse for Bayesian computation (owing to their generality and asymptotic accuracy guarantee), typically suffer data processing and computational bottlenecks as a consequence of (i) the need to hold the entire dataset (or available sufficient statistics) in memory at once; and (ii) having to evaluate of the (often expensive to compute) data likelihood at each sampling iteration. </p><p>This thesis divides into two parts. The first part concerns itself with developing efficient MCMC methods for posterior computation in the high dimensional {\em large-n large-p} setting. In particular, we develop an efficient and widely applicable approximate inference algorithm that extends MCMC to the online data setting, and separately propose a novel stochastic search sampling scheme for variable selection in high dimensional predictor settings. The second part of this thesis develops novel methods for structured sparsity in the high-dimensional {\em large-p small-n} regression setting. Here, statistical methods should scale well with the predictor dimension and be able to efficiently identify low dimensional structure so as to facilitate optimal statistical estimation in the presence of limited data. Importantly, these methods must be flexible to accommodate potentially complex relationships between the response and its associated explanatory variables. The first work proposes a nonparametric additive Gaussian process model to learn predictor-response relations that may be highly nonlinear and include numerous lower order interaction effects, possibly in different parts of the predictor space. A second work proposes a novel class of Bayesian shrinkage priors for multivariate regression with a tensor valued predictor. Dimension reduction is achieved using a low-rank additive decomposition for the latter, enabling a highly flexible and rich structure within which excellent cell-estimation and region selection may be obtained through state-of-the-art shrinkage methods. In addition, the methods developed in these works come with strong theoretical guarantees.</p> / Dissertation

This thesis presents a novel, low cost, high precision , and efficient design for an electro-hydrostatic circuit for single rod hydraulic cylinders. The design is the main contribution of candidate to fulfill the regiments of PhD degree. The challenge of existing deigns of electro-hydrostatic circuits for single-rod cylinders is using one pump to control the cylinder under switching (resistive-assistive) loads. The proposed circuit utilizes off-the-shelf industrial elements. It uses two counterbalance valves to manage switching loads and one on/off solenoid valve to redirect the differential flow of the single rod cylinder to tank. A set of simulation studies is conducted using Simhydraulic tools of Matlab in order to study performances of the proposed circuit and compare it with existing designs. Pump-controlled hydraulic circuit for double rod cylinders was developed and is widely used by industry. It is used as the benchmark for simulation studies. As well, the proposed circuit and two major existing pump-controlled circuits for single rod cylinders are compared to the benchmark circuit. Evaluations are conducted by comparing chamber pressure responses as well as pressure vs position of the cylinder end-effector for each individual circuit. Results indicate that the proposed circuit performed as well as the benchmark circuit by controlling pressures to both sides of the cylinder at the same time. Moreover, the load in the proposed circuit is more controllable compared to the benchmark circuit. Experimental results, obtained from the developed test rig, validate accuracy of the simulation model. Maximum steady state position error of 0.06 mm applications is experimentally observed when the test rig is tested under different loading conditions with various amplitudes and frequencies. The circuit consumes up to 20% of the energy that is required by a valve controlled circuit given the same sinusoidal tracking signal. The relative efficiency of the proposed circuit over a valve xii controlled circuit depends on the pattern and frequency of the tracking signal. In all the experiments, a simple proportional controller, which uses readings of a linear position transducer, is employed. The use of the proportional controller makes the proposed circuit easy to implement and shows it is good candidate for industrial applications. The accuracy of the position response of the proposed circuit indicates, it is a good candidate for robotic applications too. / May 2016

La problématique du suicide en milieu carcéral est connue et décrite dans plusieurs recherches. Toutefois les outils de dépistage du risque suicidaire ont surtout été développés pour des populations à risque non carcérales et la capacité de prédiction de ces échelles n'a été inférée qu'indirectement. Depuis un peu plus de dix ans, le questionnaire Suicide Probability Scale (Cull et Gill, 1988) est utilisée auprès de détenus québécois qui débutent une sentence fédérale. Cependant le survol de la littérature n'a pas permis de retrouver d'étude psychométrique liée à la validité prédictive de ce questionnaire auprès d'une population spécifiquement carcérale. L'objectif de la présente recherche était donc d'évaluer la valeur prédictive du questionnaire Suicide Probability Scale (SPS) auprès d'une population masculine carcérale.La recherche vise à vérifier si les détenus dépistés à risque modéré ou élevé en 1995-1996 par le SPS ont effectivement eu des comportements suicidaires par la suite, pendant qu'ils étaient encore sous la responsabilité des services correctionnels. Les résultats sont basés sur une période d'observation globale de 11 ans et demi (entre 1995 et 2006) et confirment que le SPS, dans sa forme actuelle, permet de prédire le comportement suicidaire. Une amélioration de la prédiction du risque suicidaire est démontrée si le point de découpage est modifié de 50 (point de démarcation actuel des auteurs du SPS) à 40 pour la clientèle spécifique des hommes incarcérés dans un pénitencier. Les résultats obtenus au SPS permettent aussi de dépister les détenus à risque de comportement hétéro-agressif en milieu carcéral. Le deuxième article a évalué la valeur prédictive de 24 variables actuarielles, connues en début de sentence, en combinaison avec le SPS, afin d'augmenter la prédiction et de la rendre plus spécifique en réduisant le nombre de faux négatifs.La capacité de prédiction a été analysée avec des modèles de régression logistique et la valeur sous la courbe ROC. L'ajout d'une ou de deux variables actuarielles permet d'améliorer le dépistage des comportements suicidaires sur une période de 24 mois et même de 120 mois.

近年來，電子商務飛速發展，已取得長足進步。電子商務可以高效利用有限資源，加快商業週期循環，節省時間、降低成本、提高利潤和增強企業競爭力，並以傳統經營方式所無法比擬的優點受到全球關注。因此，電子商務給企業所帶來的影響是非常重要且值得探討。 在電子商務貿易模式中，信息的高效傳遞成為平台上賣方增強競爭力的重要手段。但是過去鮮少有研究以信號理論觀點，切入探討B2B交易平台上賣方信號對買方之吸引力，以及鮮少有研究探討其中是否有其他變數可能會增加或削弱兩者之間的關係。因此本研究旨在探討賣方採用的信號與買方聯繫賣方之可能性的關係，並以企業外部市場化發展程度因素作為干擾變數。 本文共蒐集167家中小企業信息作為研究對象，並以階層回歸作為分析方法，進一步釐清賣方所採用的信號與買方聯繫賣方之可能性之間的關係，並以企業外部市場化發展程度作為干擾變數，研究外部環境情況是否會強化賣方所採用的信號與買方聯繫賣方之可能性之間的關係，補充過去文獻鮮少探討之觀點。冀望本研究結果能為平台賣方之決策者在考慮如何運營，提供一些參考與建議。

During the past decades, CEO and board compensation has increased substantially. Top management compensation and firm performance has been an extensively researched subject, and a large amount of previous studies have examined the relation of top management pay and firm performance. However, the findings and discussions have been contradictory and inconsistent. The purpose of this thesis is to examine if there is a relationship between the top management variable compensation and firm performance. We aim to explore this subject in further depth by focusing on the Swedish context and by studying if contextual issues, in terms of different industries, have an impact on this relation. In order to examine this relation multiple regression analysis were performed. The empirical evidence displays that on a general level, incentive systems of the top management have no significant effect on firm performance. We also conclude that the relation of variable pay and performance is contingent on industry. We therefore argue that the context in which the firm operates has an impact on the investigated relation in this thesis. Furthermore, the Swedish context and the Swedish governance model can be considered as one of the main explanations of the attained result.

The Pressure Impulse test-rig uses the principal energetic advantages of displacementcontrolled systems versus valve-controlled systems. The use of digital-control technology enables a high dynamic in the pressure curve, according to the requirements of ISO6605. Accumulators, along with inertia, make energy recovery possible, as well as, enabling the compression energy to be re-used. As a result of this, there is a drastic reduction in operating costs. A simulation of the system before starting the project allows the development risks to be calculated and the physically achievable performance limits to be shown.

This paper discusses the current and future drivetrain perspectives of powerful mobile machines, especially in regards to TCO and drive performance. For the TCO-impact, the power losses of the components plays a big role and, if they are designed for efficiency, they have a significant and measurable influence. From the braking function point of view, this paper demonstrates not only the advantages of a valve-based over a control algorithm based solution, but also its innovative development directions towards a more sophisticated engine speed controller with optimized heat conversion into the oil. Also for the drivetrain subsystems, innovative components are discussed, like the hybrid control, combining the benefits of a non-feedback and a displacement control in one single assembly, or the variable charge system for further reduced energy consumption of the overall drivetrain.

The Gravity Recovery and Climate Experiment (GRACE) is a joint National Aeronautics and Space Administration / Deutsches Zentrum für Luftund Raumfahrt (NASA/DLR) mission to map the time-variable and mean gravity field of the Earth, and was launched on March 17, 2002. The nature of the gravity field inverse problem amplifies the noise in the data that creeps into the mid and high degree and order harmonic coefficients of the earth's gravity fields for monthly variability, making the GRACE estimation problem ill-posed. These errors, due to the use of imperfect models and data noise, are manifested as peculiar errors in the gravity estimates as north-south striping in the monthly global maps of equivalent water heights. In order to reduce these errors, this study develops a methodology based on Tikhonov regularization technique using the L-curve method in combination with orthogonal transformation method. L-curve is a popular aid for determining a suitable value of the regularization parameter when solving linear discrete ill-posed problems using Tikhonov regularization. However, the computational effort required to determine the L-curve can be prohibitive for a large scale problem like GRACE. This study implements a parameter-choice method, using Lanczos bidiagonalization that is a computationally inexpensive approximation to L-curve called L-ribbon. This method projects a large estimation problem on a problem of size of about two orders of magnitude smaller. Using the knowledge of the characteristics of the systematic errors in the GRACE solutions, this study designs a new regularization matrix that reduces the systematic errors without attenuating the signal. The regularization matrix provides a constraint on the geopotential coefficients as a function of its degree and order. The regularization algorithms are implemented in a parallel computing environment for this study. A five year time-series of the candidate regularized solutions show markedly reduced systematic errors without any reduction in the variability signal compared to the unconstrained solutions. The variability signals in the regularized series show good agreement with the hydrological models in the small and medium sized river basins and also show non-seasonal signals in the oceans without the need for post-processing. / text

Impedance measurement as a function of frequency is being increasingly used for the detection of organic molecules. The main building block required for this is a sinusoidal oscillator whose frequency can be varied in the range of a few KHz to tens of MHz. The thesis describes the design of Integrated CMOS Oscillator Circuits. There are 2 designs presented in the thesis, one of which is based on the Wien Bridge and the other, on an LC architecture. They provide both in-phase and quadrature outputs needed for the determination of the real and imaginary parts of complex impedances. The inductor in the LC tank is realized by gyration of a capacitor. This needs two variable transconductance elements. Linear transconductance elements with decoupled transconductance gm and output conductance go is presented. A novel circuit for detecting and controlling the amplitude of oscillation is described. A current mode technique to scale the capacitance is also discussed. Since this oscillator is used in an inexpensive hand-held instrument, both power consumption and chip area must be minimized. A comparison between the Wien Bridge and the LC tank based oscillator is presented. Simulation results pertaining to the design of the different blocks of the circuit are made available. / text