京都大学 / 新制・課程博士 / 博士(工学) / 甲第24613号 / 工博第5119号 / 新制||工||1979(附属図書館) / 京都大学大学院工学研究科原子核工学専攻 / (主査)教授 村上 定義, 教授 横峯 健彦, 教授 宮寺 隆之 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

Past political regimes and socio-economic imbalances have led to the formation of a transport system in the Republic of South Africa (RSA) that is unique to the developing world. Affluent communities in metropolitan cities are situated close to economic activity, whereas the people in need of public transport are situated on the periphery of the cities. This demographic structure is opposite to that of developed countries and complicates both the provision of transport services and the planning process thereof. Multi-Agent Transport Simulation (MATSim) has been identified as an Agent-Based Simulation (ABS) approach that models individual travellers as autonomous entities to create large scale traffic simulations. The initial implementation of MATSim in the RSA successfully simulated private vehicle trips between home and work in the province of Gauteng, proving that there is enough data available to create a realistic multi-agent transport model. The initial implementation can be expanded to further enhance the simulation accuracy, but this requires the incorporation of additional primary and secondary activities into the initial transport demand. This study created a methodology to expand the initial implementation in the midst of limited data, and implemented this process for Gauteng. The first phase constructed a 10% synthetic population that represents the demographic structure of the actual population and identified various socio-demographic attributes that can influence an individual's travel behaviour. These attributes were assigned to the synthetic agents by following an approach that combines probabilistic sampling and rule-based models. The second phase used agents' individual attributes, and census, National Household Travel Survey (NHTS) and geospatial data to transform the synthetic population into a set of daily activity plans - one for every agent. All the agents' daily plans were combined into a plans.xml file that was used as input to MATSim, where the individuals' activity plans were executed simultaneously to model the transport decisions and behaviour of agents. Data deficiencies were overcome by contemplating various scenarios and comparing the macroscopic transport demand patterns thereof to the results of the initial implementation and to actual counting station statistics. This study successfully expanded the initial home-work-home implementation of MATSim by including additional non-work activities in the transport demand. The addition of non-work activities improved the simulation accuracy during both peak and off-peak periods, and the initial demand therefore provides an improved representation of the travel behaviour of individuals in Gauteng. / Dissertation (MEng)--University of Pretoria, 2011. / Industrial and Systems Engineering / unrestricted

Title: Adjoint-Driven Importance Sampling in Light Transport Simulation Author: RNDr. Jiří Vorba Department: Department of Software and Computer Science Education Supervisor: doc. Ing. Jaroslav Křivánek, Ph.D., Department of Software and Computer Science Education Abstract: Monte Carlo light transport simulation has recently been adopted by the movie industry as a standard tool for producing photo realistic imagery. As the industry pushes current technologies to the very edge of their possibilities, the unprecedented complexity of rendered scenes has underlined a fundamental weakness of MC light transport simulation: slow convergence in the presence of indirect illumination. The culprit of this poor behaviour is that the sam- pling schemes used in the state-of-the-art MC transport algorithms usually do not adapt to the conditions of rendered scenes. We base our work on the ob- servation that the vast amount of samples needed by these algorithms forms an abundant source of information that can be used to derive superior sampling strategies, tailored for a given scene. In the first part of this thesis, we adapt general machine learning techniques to train directional distributions for biasing scattering directions of camera paths towards incident illumination (radiance). Our approach allows progressive...

This dissertation presents the design of an integrated watershed model, WASH123D version 3.0, a first principle, physics-based watershed-scale model of integrated hydrology/hydraulics and water quality transport. This numerical model is comprised of three modules: (1) a one-dimensional (1-D) simulation module that is capable of simulating separated and coupled fluid flow, sediment transport and reaction-based water quality transport in river/stream/canal networks and through control structures; (2) a two-dimensional (2-D) simulation module, capable of simulating separated and coupled fluid flow, sediment transport, and reactive biogeochemical transport and transformation in two-dimensional overland flow systems; and (3) a three-dimensional (3-D) simulation module, capable of simulating separated and coupled fluid flow and reactive geochemical transport and transformation in three-dimensional variably saturated subsurface systems. The Saint Venant equation and its simplified versions, diffusion wave and kinematic wave forms, are employed for surface fluid flow simulations and the modified Richards equation is applied for subsurface flow simulation. The reaction-based advection-dispersion equation is used as the governing equation for water quality transport. Several physically and mathematically based numerical options are provided to solve these governing equations for different application purposes. The surface-subsurface water interactions are considered in the flow module and simulated on the basis of continuity of interface. In the transport simulations, fast/equilibrium reactions are decoupled from slow/kinetic reactions by the decomposition of reaction networks; this enables robust numerical integrations of the governing equation. Kinetic variables are adopted as primary dependent variables rather than biogeochemical species to reduce the number of transport equations and simplify the reaction terms. In each time step, hydrologic/hydraulic variables are solved in the flow module; kinetic variables are then solved in the transport module. This is followed by solving the reactive chemical system node by node to yield concentrations of all species. Application examples are presented to demonstrate the design capability of the model. This model may be of interest to environmental scientists, engineers and decision makers as a comprehensive assessment tool to reliably predict the fluid flow as well as sediment and contaminant transport on watershed scales so as to evaluate the efficacy and impact of alternative watershed management and remediation techniques prior to incurring expense in the field.

Light transport simulation is used in realistic image synthesis to create physically plausible images of virtual scenes. Important components of the scenes are participating media (e.g. air, water, skin etc.). Efficient computation of light transport in participating media robust to their large diversity is still an open problem. We implemented the UPBP algorithm recently developed by Křivánek et al. It addresses the problem by combining several complementary previous methods using multiple importance sampling, and excels at rendering scenes where the previous methods alone fail. The implementation is available online, we focused on its thorough description to facilitate and support further research in this field. Powered by TCPDF (www.tcpdf.org)

Le transistor est la brique élémentaire des circuits intégrés présents dans tous les appareils électroniques. Années après années l’industrie de la microélectronique a amélioré les performances des circuits intégrés (rapidité, consommation énergétique) en réduisant les dimensions du transistor. De nos jours, en plus de la réduction de la taille du transistor d’autres techniques permettent de soutenir cette croissance: ce sont les « booster » technologiques. Les contraintes mécaniques ou encore le remplacement du Silicium par d’autres matériaux tels que germanium (Ge) et les matériaux semi-conducteurs de type III-V sont des exemples de booster technologiques. Grâce à la modélisation numérique, cette thèse propose d’étudier les effets de booster technologiques sur les performances électriques de la future génération de transistors. / The transistor is the elementary brick of Integrated circuits found in all electronic devices. Years after years the microelectronic industry has enhanced the performances of integrated circuits (speed and energy consumption) by downscaling the transistor. Nowadays besides the transistor’s downscaling, other techniques have been considered to maintain this growth: they are called technological boosters. Mechanical strain or new material, such as germanium (Ge) and III-V semiconductors, to replace Silicon are example of technological boosters. By the means of numerical quantum simulations and modeling, this these work propose a study of the effect of technological boosters on the electric performances of the next generation of transistors.

Monte Carlo light transport simulation has become a de-facto standard tool for photorealistic rendering. However, the algorithms used by the current rendering systems are often ineffective, especially in scenes featuring light transport due to multiple highly glossy or specular interactions and complex visibility between the camera and light sources. It is therefore desirable to adopt more robust algorithms in practice. Light transport algorithms based on Markov chain Monte Carlo (MCMC) are known to be effective at sampling many different kinds of light transport paths even in the presence of complex visibility. However, the current MCMC algorithms often over-sample some of the paths while under-sampling or completely missing other paths. We attribute this behavior to insufficient global exploration of path space which leads to their unpredictable convergence and causes the occurrence of image artifacts. This in turn prohibits adoption of MCMC algorithms in practice. In this thesis we therefore focus on improving global exploration in MCMC algorithms for light transport simulation. First, we present a new MCMC algorithm that utilizes replica exchange to improve global exploration. To maximize efficiency of replica exchange we introduce tempering of the path space, which allows easier discovery of important...

Not available / text

Not available / text