We consider the motion of a Brownian particle trapped in an arbitrary bounded two or three-dimensional domain, whose boundary is reflecting except for a small absorbing window through which the particle can escape. We use the method of matched asymptotic expansions to calculate the mean first passage time, defined as the time taken for the Brownian particle to escape from the domain through the absorbing window. This is known as the narrow escape problem. Since the mean escape time diverges as the window shrinks, the calculation is a singular perturbation problem. We extend our results to include N absorbing windows of varying length in two dimensions and varying radius in three dimensions. We present findings in two dimensions for the unit disk, unit square and ellipse and in three dimensions for the unit sphere. The narrow escape problem has various applications in many fields including finance, biology, and statistical mechanics. / Science, Faculty of / Mathematics, Department of / Graduate

Automated narrow-gap girth-butt welds are replacing conventional welding methods to join sections of austenitic stainless steel pipe in the primary circuit of Pressurised Water Reactors, to reduce manufacturing costs and improve quality. To ensure the safe operation of these systems, reliable structural integrity assessments have to be undertaken, requiring the mechanical properties of welded joints to be characterised alongside the weld residual stress magnitude and distribution.This research project characterised, for the first time, the weld residual stress field and the tensile and ductile fracture toughness properties of a 33 mm thick narrow-gap 304L stainless steel pipe weld. The residual stress was characterised using two complementary approaches: deep hole drilling and neutron diffraction. A novel neutron diffraction scanning technique was developed to characterise the residual stress field, without cutting an access window into the component, leaving the original weld residual stress field undisturbed. A modified deep hole drilling technique was developed to characterise the residual stress retained in fracture mechanics specimens extracted from the pipe weld in two orientations. The modified technique was shown to measure the original weld residual stress field more accurately than through conventional deep hole drilling. Residual stresses, exceeding 50% of the weld material proof strength, were retained in axially-orientated fracture mechanics specimens.Tensile tests showed that the weld was approximately 60% overmatched. It was demonstrated that neither retained residual stress, nor specimen orientation, had a discernible effect on the measured fracture toughness of the weld material. In less ductile materials, however, the level of retained residual stress may unduly influence the measurement of fracture toughness. At initiation, the fracture toughness properties of both the parent and weld materials were far in excess of the measuring capacity of the largest fracture mechanics specimens that could be machined from the weld.The influence of residual stress and fracture toughness on the performance of narrow-gap welded pipework was investigated. Full elastic-plastic finite element analyses were used to model the pipe weld, containing a postulated defect under combined primary and secondary loading. The results, applied within the framework of an R6 structural integrity assessment, compared different plasticity interaction parameters on the prediction of failure load; the conventional ρ-parameter approach was compared with the recently developed, more advanced, g-parameter. It was shown that the g-parameter significantly reduced the conservatism of the ρ-parameter approach. However, for this pipe weld, plastic collapse was predicted to precede failure by ductile initiation, suggesting that a plastic collapse solution may be an appropriate failure criterion to use in structural integrity assessments of similar component and defect combinations.

This paper proposes a numerical finite-element-based model aimed at optimizing narrow-vein stope stability. This model combines empirical and numerical methods to develop a sequence, which may determine an acceptable stope safety factor. A stope stability analysis was conducted through the Mathews stability graph method, which requires two factors: the hydraulic radius (HR) and stability number (N'). The Mathews stability graph method is used to assess the stability of an underground design. Variations in stope dimensions are estimated by changing the HR and Factor A within the N', which is determined through numerical methods. The results of the numerical simulation indicate that the HR increases with an increase in stope dimensions, while Factor A maintains an inverse relationship with the maximum stress induced on the excavation walls. This document demonstrates the potential of combining empirical and numerical methods in stope design optimization, especially when developed in small narrow vein mines.

The quenching of undesired flames by cold surfaces has been investigated for more than a century. The current quenching theory can predict simple configurations, this is not the case for real environments such as fuel management systems. Flames are sensitive to numerous parameters, such as fuel, mixture fraction, pressure, temperature, flow properties, acoustics, radiation, and surface interactions. The effects of some of these parameters are very well documented but there is a lack of information regarding the effects of acoustics and flow. This dissertation work will focus on improving the understanding of flow effect on the quenching of premixed gaseous flames. First, the effect of apparent velocity on flame quenching was investigated for different fuels and equivalence ratios. An experimental facility is designed such that the apparent flame velocity at which the flame enters and propagates through the channel can be varied without changing the initial mixture condition. High-speed (15,000 frames per second (FPS)) Schlieren and dynamic pressure measurement were used to measure the apparent flame velocity and to assess the flame quenching, respectively. This study showed that the high-speed laminar flames are harder to quench compared to self-propagating and turbulent flames. A similar trend was obtained for all the conditions investigated, lean and stoichiometric methane-air, lean propane-air, and lean ethylene-air mixtures. Further investigation was carried out to understand the quenching of high-speed laminar flames. The flame propagation through the channel was investigated using Hydroxyl (OH) planar laser induced fluorescence (PLIF). This study showed that the OH intensity fell below the detection threshold in the later part of the channel when quenching is observed. Then, the influence of heat transfer was investigated using spatial and temporal evolution of the temperature in the quenching channel. A high-speed (10 kHz) filtered Rayleigh scattering (FRS) technique was used to measure the one-dimensional time-resolved temperature profile. Three different channel heights (H = 1.3, 1.5, 2.0 mm) were investigated. Based on the evolution of the temperature profile in the quenching channel, a new parameter was identified and the importance of its evolution on the flame quenching was discussed.

The objective of this thesis is to study the demodulation of narrowband radio frequency signals by aliasing sampling in order to reduce the sampling rate. The spectrum can be recreated at the lower frequency position by aliasing sampling. However, if the sampling rate is deviated from the desired one, error will occur. The sensitivity to the frequency error of aliasing sampling is studied. One main reason of the deviation of the sampling rate is the frequency drifting of the local oscillator. Being able to compensate the oscillator drifting errors inexpensively, automatic frequency control (AFC) loops are important at receivers. Two major digital AFC algorithms are studied. One is the Phase method AFC, and the other is the Magnitude method AFC. Study indicates that both methods perform almost equally well. One adaptive AFC algorithm is also proposed. The scheme of the adaptive AFC algorithm is to use Upper-bound and Lower-bound techniques to squeeze the frequency errors. It is shown that the adaptive AFC algorithm can achieve up to 20 dB average signal-to-noise power ratio over the Magnitude method AFC under a noisy environment.

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

Den här uppsatsen har som syfte att undersöka huruvida det är möjligt att ta fram riktlinjer för vad privata verksamheter behöver ta i beaktande inför en planerad implementering av artificiell intelligens. Studien kommer belysa faktorer som hjälper företag att förstå vad som krävs inför en sådan omställning, men även de hinder som måste övervinnas för att lyckas. Studiens datainsamling har genomförts med två metoder, först en litteraturstudie sedan kvalitativa, semistrukturerade forskningsintervjuer. Dessa har sedan analyserats med vars en analysmetod som kompletterar varandra och därefter tolkats för att se mönster som kan besvara studiens frågeställning: Vad måste svenska organisationer inom den privata sektorn beakta för att lyckas implementera Artificiell intelligens i sin verksamhet? Resultatet har tagits fram genom att jämföra vetenskapliga texter och intervjuer, för att undersöka om den akademiska och praktiska synen skiljer sig åt. Studien resulterade i åtta faktorer som företag borde ta i beaktning inför en implementering av artificiell intelligens. Författarna hoppas att med den här studien kunna främja svensk utveckling inom artificiell intelligens och på så vis generera ett större nationellt mervärde och en starkare konkurrenskraft internationell. / The purpose of this paper is to investigate the possibilities to develop guidelines for businesses to take into account before an implementation of artificial intelligence. The study will highlight different factors that will help companies to understand what is required to make this kind of digital transition, it will also highlight the obstacles companies have to overcome in order to succeed. The data collection was conducted in two parts, first a literature study and then qualitative, semi-structured interviews. These were analyzed with their own analysis which supplement each other, and interpreted to identify patterns that could answer the study's main question: What must Swedish organizations in the private sector consider in order to successfully implement Artificial Intelligence in their operations?The result of the study has been produced by comparing scientific texts and interviews, to investigate whether the academic and practical views differ. The study resulted in eight factors that companies should consider before implementing artificial intelligence. The authors hope that the study will promote Swedish development in artificial intelligence and thus generate a greater national value and international competitiveness.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第19522号 / 理博第4182号 / 新制||理||1600(附属図書館) / 32558 / 京都大学大学院理学研究科化学専攻 / (主査)教授 吉村 一良, 教授 有賀 哲也, 教授 北川 宏 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第21353号 / 工博第4512号 / 新制||工||1703(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 藤田 正治, 教授 中川 一, 准教授 竹林 洋史 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Personality, as a frequently used predictor of job performance, has often been criticized for its low criterion-related validity when compared to cognitive tests and some other predictors. The present study investigated incremental validity of narrow-trait personality to distinguish predictive from non-predictive facets of Conscientiousness and Extraversion dimensions. In addition, some intermediate mechanisms that may link the two personality dimensions with the criterion, such as different types of person-environment (P-E) fit and job involvement variables, were tested as well. The institution's job performance scale, NEO-PI-3 personality scale, person-organization (P-O) fit, needs-supplies (N-S) fit, demands-abilities (D-A) fit, job involvement questionnaire (JIQ), and demographic measures were administered in an online survey to 295 professional and civil service employees of a midsize Midwestern university. The sample was predominantly female and Caucasian with a mean age of 45.8 years and a median length of current employment of 5.1 years. Both personality dimensions were positively related to overall job performance. Conscientiousness was a stronger predictor of task performance, whereas Extraversion was related more consistently to contextual performance. In stepwise multiple regression analyses containing facets of personality dimensions as predictors of overall job performance, Competence emerged as the only facet of Conscientiousness, and Warmth and Assertiveness as the only facets of Extraversion that accounted for a meaningful amount of variance in the criterion. The use of narrow-trait personality to predict overall job performance enhances criterion-related validity of the construct and renders it a more efficient predictor of job performance than global-trait personality. Among the potential mediators, P-O and D-A fit partially mediated the personality-performance relationship providing evidence for the importance of perceptions of congruence in values and the ability to meet demands of the job. Current results are considered in light of limitations. Implications for theory, research, and practice, as well as future research directions are discussed.