本論文共一冊，章節內容如左： 第一章：導論（主要在說明研究的背景、動機；研究問題與研究目的、研究範圍、限 制）。 第二章：理論架構與文獻探討（先說明變數（企業、市場、競爭變數）的構面及相關 文獻，再探討企業策略分類與績效的相關文獻）。 第三章：研究方法（先說明研究變數的操作性定義，及測量，再寫出研究架構、假設 與限制。） 第四章：研究發現（先說明影響績效的主要策略及變數，再探討它們的直接與間接效 果）。 第五章：結論與建議（廠商已擁有何種優勢時，應採何種策略才可提高績效，後續者 的研究方向）。

Nous étudions la production de couple et de polluants de moteurs à combustion interne à distribution variable d'un type fréquemment répandu dans l'industrie automobile (moteurs Diesel ou essence). Les systèmes de distribution variable (VVA pour Variable Valve Actuation) ont été introduits dans les moteurs pour optimiser l'efficacité du fonctionnement global sur des plages très larges de points de fonctionnement, considérés à l'équilibre. A tout instant, les résultats de la combustion dépendent du remplissage en air/gaz brûlés du cylindre, qui lui même dépend de la position des actionneurs VVA et des conditions dans le collecteur d'admission. Ces deux sous-systèmes ont des temps de réponse différents, ce qui conduit, pendant les transitoires, à des pertes d'efficacit ́e du moteur. Dans cette thèse, nous proposons une solution a ce problème, qui prend la forme de boucle de coordination entre les sous-systèmes de contrôle de bas-niveau. Cette coordination s'effectue par l'intermédiaire d'un modèle de remplissage développé dans ce manuscrit. Des résultats expérimentaux établissent qu'il est possible d'améliorer la production de couple et de réduire les polluants.

In this work, we present a way to extend Ant Colony Optimization (ACO), so that it can be applied to both continuous and mixed-variable optimization problems. We demonstrate, first, how ACO may be extended to continuous domains. We describe the algorithm proposed, discuss the different design decisions made, and we position it among other metaheuristics. Following this, we present the results of numerous simulations and testing. We compare the results obtained by the proposed algorithm on typical benchmark problems with those obtained by other methods used for tackling continuous optimization problems in the literature. Finally, we investigate how our algorithm performs on a real-world problem coming from the medical field—we use our algorithm for training neural network used for pattern classification in disease recognition. Following an extensive analysis of the performance of ACO extended to continuous domains, we present how it may be further adapted to handle both continuous and discrete variables simultaneously. We thus introduce the first native mixed-variable version of an ACO algorithm. Then, we analyze and compare the performance of both continuous and mixed-variable ACO algorithms on different benchmark problems from the literature. Through the research performed, we gain some insight into the relationship between the formulation of mixed-variable problems, and the best methods to tackle them. Furthermore, we demonstrate that the performance of ACO on various real-world mixed-variable optimization problems coming from the mechanical engineering field is comparable to the state of the art.

Ce travail a comme objectif l'amélioration et l'extension de l'application des systèmes de réalité virtuelle en retour haptique dans le cadre d'un travail collaboratif à distance induisant par conséquent des retards de transmission. Les développements des lois de commande présentées dans ce document font partie de celles améliorant les schémas de commande bilatérale classiques. Sous cette bannière, on trouve deux catégories de commande : passives, comme celles basées sur les variables d'onde où prédictives, en se basant sur l'estimation du modèle de l'esclave.

.

In this work we investigate the use of perfect sampling methods within the context of Bayesian linear regression. We focus on inference problems related to the marginal posterior model probabilities. Model averaged inference for the response and Bayesian variable selection are considered. Perfect sampling is an alternate form of Markov chain Monte Carlo that generates exact sample points from the posterior of interest. This approach removes the need for burn-in assessment faced by traditional MCMC methods. For model averaged inference, we find the monotone Gibbs coupling from the past (CFTP) algorithm is the preferred choice. This requires the predictor matrix be orthogonal, preventing variable selection, but allowing model averaging for prediction of the response. Exploring choices of priors for the parameters in the Bayesian linear model, we investigate sufficiency for monotonicity assuming Gaussian errors. We discover that a number of other sufficient conditions exist, besides an orthogonal predictor matrix, for the construction of a monotone Gibbs Markov chain. Requiring an orthogonal predictor matrix, we investigate new methods of orthogonalizing the original predictor matrix. We find that a new method using the modified Gram-Schmidt orthogonalization procedure performs comparably with existing transformation methods, such as generalized principal components. Accounting for the effect of using an orthogonal predictor matrix, we discover that inference using model averaging for in-sample prediction of the response is comparable between the original and orthogonal predictor matrix. The Gibbs sampler is then investigated for sampling when using the original predictor matrix and the orthogonal predictor matrix. We find that a hybrid method, using a standard Gibbs sampler on the orthogonal space in conjunction with the monotone CFTP Gibbs sampler, provides the fastest computation and convergence to the posterior distribution. We conclude the hybrid approach should be used when the monotone Gibbs CFTP sampler becomes impractical, due to large backwards coupling times. We demonstrate large backwards coupling times occur when the sample size is close to the number of predictors, or when hyper-parameter choices increase model competition. The monotone Gibbs CFTP sampler should be taken advantage of when the backwards coupling time is small. For the problem of variable selection we turn to the exact version of the independent Metropolis-Hastings (IMH) algorithm. We reiterate the notion that the exact IMH sampler is redundant, being a needlessly complicated rejection sampler. We then determine a rejection sampler is feasible for variable selection when the sample size is close to the number of predictors and using Zellner’s prior with a small value for the hyper-parameter c. Finally, we use the example of simulating from the posterior of c conditional on a model to demonstrate how the use of an exact IMH view-point clarifies how the rejection sampler can be adapted to improve efficiency.

Hydrologic catchment response varies in time. The goal of this dissertation is to investigate how and why it varies and what controls these variations. In order to tackle these questions the first step is to develop a method that permits the capturing of the temporal variation of transit time distributions (TTDs). To this end, the established transfer function-convolution approach using time series of stable water isotopes was modified so that it is now able to determine variable mean transit times (mTTs). The type and the shape parameter of the transfer function also vary in time. We found that antecedent moisture content, saturated hydraulic conductivity, soil depth and subsequent precipitation intensity are all potential controls. We propose a dimensionless number that integrates these controls and relates available storage to incoming and outgoing water fluxes in combination with information on antecedent moisture conditions to predict TTD type and shape. The individual TTDs for every time step produced by this model can be superimposed, summed and normalized to create a classification tool for catchments that is based on its general response behavior to precipitation events: the master transit time distribution. With this model in hand the hydrologic response for three consecutive monsoon seasons in ten nested subcatchments was examined. It was found that the major response controls were changing between the years in accordance with three hydrologic response modes. The mTT correlated most strongly with soil depth in the first year, with hydraulic conductivity in the second year and with curvature in the third year. These variations were produced by differences in precipitation patterns that led to differences in soil saturation and consequently to different dominant flow processes: in the first year most of the water left the catchment via fast flow paths (macropore flow, overland flow), in the second year shallow subsurface flow in the soil matrix was more dominant and in the third year most outflowing water derived from slow base flow. To better predict hydrologic catchment response we propose to apply a dimensionless number to determine the catchment response mode for every time step before selecting the appropriate response control.

Great portion of today's industry are <em>Adjustable Speed Drives</em> (<em>ASD's</em>) operated in order to fulfill certain processes. When these processes are critical ones or sensitive to voltage disturbances, that might take place due to inserting high load in an area near to the Point of Common Coupling (<em>PCC</em>) of the process or due to a short term outage, few tens of thousands up to millions of dollars will be lost once such interruptions (voltage sags) take place as a result of the process failure. On the other hand, a distorted voltage waveform at the PCC for some sensitive process might malfunction as a result of the high harmonic content of the voltage waveform. Utilities are required to deliver as pure as possible sinusoidal voltage waveform according to certain limits; thus, they might apply fines against the consumers who are responsible for producing high amounts of current harmonics that affect the voltage wave shape at the <em>PCC</em> in order to force them to improve the consumer's load profile by adding filters at <em>PCC</em> for instance. Utilities are charging the consumers who are drawing power at poor power factor as well. <br /><br /> This thesis presents an <em>ASD</em> retrofitted with a dual-functional piece of hardware connected in series to its <em>DC-link</em> that is capable of handling the previously two mentioned problems. In other words, hardware that is capable of providing <em>voltage sag ride-through</em> during the voltage sag conditions on one side, on the other side, during the normal operating conditions, it is capable to mitigate the harmonic contents of the drawn current by the ASD's rectifier and to improve the power factor. <br /><br /> Survey on voltage sag ride-through for ASD's approaches are presented in the literature has been made. Approaches are classified as the topology utilized; first, topologies that utilizes energy storage elements that store energy to compensate the DC-link voltage with during the voltage sags, second, topologies retrofitting the DC-link itself with additional hardware to compensate the DC-link voltage. The first group is capable to provide voltage compensating during the full outages while the second can't. The presented voltage sag ride-through work of this thesis belongs to the second group. <br /><br /> Boost converter has been used as the hardware to compensate the DC-link voltage because of its simplicity and cheap price. An adaptive linear network (ADALINE) is investigated as the detection system to detect the envelope of the input voltage waveform. Once the envelope of the voltage goes below a certain level, the boost converter is activated to compensate the difference between voltage set point and the actual DC-link voltage. Simulation results supporting the proposed configuration are presented. <br /><br /> A third-harmonic current injection approach is utilized in this work in order to achieve <em>total harmonic distortion</em> (<em>THD</em>) mitigation from 32% to 5. 125% (theoretically). Two third-harmonic current injection networks have been investigated; one utilizes a real resistor, the other utilizes a resistor emulator to reduce the energy dissipated. The proposed controller for the resistor emulator does not require a proportional-integral (PI) controller. <br /><br /> As a result of the common devices between the voltage sag ride-through circuitry and the harmonic mitigation one, they can be integrated together in one circuitry connected in series with the DC-link of the ASD. And hence, the dual functionality of the hardware will be achieved. Simulation results supporting the theoretical results have been presented.

Adhesive joining is an attractive alternative to conventional joining methods, such as welding and mechanical fastening. The benefits of adhesive bonding include: the ability to form lightweight, high stiffness structures; joining of different types of materials; better fatigue performance, and reduction in the stress concentrations or the effects of the heat associated with welding. However, concerns about the durability of adhesive joints still hinder their widespread use in structural applications. Moisture has been identified as one of the major factors affecting joint durability. This is especially important in applications where joints are exposed to varying moisture conditions throughout their useful life. The aim of this research is to develop models to predict degradation in adhesive joints under varying moisture conditions. This was achieved by a combination of experimental and numerical methods. Experiments were carried out to characterise the moisture uptake and mechanical properties of the single part epoxide adhesive, FM73-M. Single lap joints were manufactured from aluminium alloy 2024 in heat treated (T3) and non heat treated (O) states using the FM73-M, BR127 adhesive-primer system. Tensile testing of the single lap joints was carried out after the joints had been exposed to hot-wet conditioning environments. Models were developed for predicting moisture concentration in the adhesive under cyclic moisture absorption and desorption conditions. A finite element based methodology incorporating moisture history was developed to predict the cyclic moisture concentration. In the next step, a novel finite element based methodology, which was based on moisture history effects, was developed to determine stresses in bonded joints after curing, conditioning and tensile testing. In the final step, a moisture history dependent cohesive zone element based damage and failure criterion was introduced to predict damage initiation, crack growth and failure under variable moisture and temperature conditions. The methodology proposed in this work and its implementation by finite element method provides a systematic approach for determining the degradation in adhesive joints under varying environmental conditions and accomplishes the aim of this research.

This research investigates a batch arrival queueing system with a Bernoulli scheduled vacation and random system breakdowns. It is assumed that the repair process does not start immediately after the breakdown. Consequently there maybe a delay in starting repairs. After every service completion the server may go on an optional vacation. When the original vacation is completed the server has the option to go on an extended vacation. It is assumed that the system is equipped with a stand-by server to serve the customers during the vacation period of the main server as well as during the repair process. The service times, vacation times, repair times, delay times and extended vacation times are assumed to follow different general distributions while the breakdown times and the service times of the stand-by server follow an exponential distribution. By introducing a supplementary variable we are able to obtain steady state results in an explicit closed form in terms of the probability generating functions. Some important performance measures including; the average length of the queue, the average number of customers in the system, the mean response time, and the value of the traffic intensity are presented. The professional MathCad 2001 software has been used to illustrate the numerical results in this study.