No description available.

Thesis (doctoral)--Lund University, 1992. / Added t.p. with thesis statement inserted.

Thesis (M. Ed.)--University of Hong Kong, 1996. / Includes bibliographical references (leaf 135-140).

Thesis (doctoral)--Lund University, 1992. / Added t.p. with thesis statement inserted.

Thesis (M.Ed.)--University of Hong Kong, 1996. / Includes bibliographical references (leaves 135-140). Also available in print.

Modéliser le mouvement des foules humaines est essentiel pour des situations aussi diverses que la prévention de risque dans les lieux publics, la planification d’évènements ou la création d’animations visuelles réalistes. Cependant, la difficulté de mener des expériences quantitatives limite notre compréhension de la dynamique des piétons, et le manque de mesures de référence rend impossible une comparaison poussée des modèles existants. Cette thèse tente d’augmenter notre compréhension des foules humaines par deux approches distinctes. Dans un premier temps, nous avons conduit une étude numérique et théorique pour étudier formation de lignes au sein de flux bidirectionnels d'agents motiles. Nous avons montré qu’une transition de phase critique du second ordre séparait un état mélangé d’un état constitué de lignes géantes le long desquelles se déplacent les agents visants une même direction. Cette séparation est caractéristique des systèmes actifs. Une approche hydrodynamique nous a ensuite permis de prouver que les phases mélangées sont aussi algébriquement corrélées dans la direction longitudinale. Nous avons expliqué et montré que ces fortes corrélations sont génériques de tous systèmes de flux bidirectionnels, qu’ils soient constitués de particules forcées ou de particules actives. Dans un second temps, nous avons mené une campagne expérimentale de grande envergure afin d’établir une expérience de référence des foules humaines. Nous avons pour cela choisi un système modèle, la zone d’attente de marathons. Dans ces foules de dizaines de milliers d’individus, nous avons quantitativement établi que les fluctuations de vitesse se propagent sur de grandes échelles, alors que les variations d’orientation s’évanouissent en quelques secondes. Grâce à ces mesures, nous avons construit une théorie prédictive hydrodynamique des foules polarisées. / Modelling crowd motion is central to situations as diverse as risk prevention in mass events and visual effects rendering in the motion picture industry. The difficulty to perform quantitative measurements in model experiments, and the lack of reference experimental system, have however strongly limited our ability to model and control pedestrian flows. The aim of this thesis is to strengthen our understanding of human crowds, following two distinct approaches.First, we designed a numerical model to study the lane formation process among bidirectional flows of motile particles. We first evidenced the existence of two distinct phases: one fully laned and one homogeneously mixed, separated by a critical phase transition, unique to active systems. We then showed with a hydrodynamic approach that the mixed phase is algebraically correlated in the direction of the flow. We elucidated the origin of these strong correlations and proved that they were a universal feature of any system of oppositely moving particles, active of passive.Second, we conducted a substantial experimental campaign to establish a model experiment of human crowds. For that purpose we performed systematic measurements on crowds composed of tens of thousands of road-race participants in start corrals, a geometrically simple setup. We established that speed information propagates through polarized crowds over system spanning scales, while orientational information is lost in a few seconds. Building on these observations, we laid out a hydrodynamic theory of polarized crowds and demonstrated its predictive power.

This thesis describes a proposed design and implementation of a predictive maintenance engine developed to fulfill the requirements of the STO Company (Societe de transport de l'Outaouais) for maintaining vehicles in the fleet. Predictive maintenance is proven to be an effective approach and has become an industry standard in many fields. However, in the transportation industry, it is still in the stages of development due to the complexity of moving systems and the high level dimensions of involved parameters. Because it is almost impossible to cover all use cases of the vehicle operational process using one particular approach to predictive maintenance, in our work we take a systematic approach to designing a predictive maintenance system in several steps. Each step is implemented at the corresponding development stage based on the available data accumulated during system funсtioning cycle. % by dividing the entire system into modules and implementing different approaches. This thesis delves into the process of designing the general infrastructural model of the fleet management system (FMS), while focusing on the edge gateway module located on the vehicle and its function of detecting maintenance events based on current vehicle status. Several approaches may be used to detect maintenance events, such as a machine learning approach or an expert system-based approach. While the final version of fleet management system will use a hybrid approach, in this thesis paper we chose to focus on the second option based on expert knowledge, while machine learning has been left for future implementation since it requires extensive training data to be gathered prior to conducting experiments and actualizing operations. Inspired by the IDEA methodology which promotes mapping business rules as software classes and using the object-relational model for mapping objects to database entities, we take active database features as a base for developing a rule engine implementation. However, in contrast to the IDEA methodology which seeks to describe the specific system and its sub-modules, then build active rules based on the interaction between sub-systems, we are not aware of the functional structure of the vehicle due to its complexity. Instead, we develop a framework for creating specific active rules based on abstract classifications structured as ECA rules (event-condition-action), but with some expansions made due to the specifics of vehicle maintenance. The thesis describes an attempt to implement such a framework, and particularly the rule engine module, using active database features making it possible to encapsulate the active behaviour inside the database and decouple event detection from other functionalities. We provide the system with a set of example rules and then conduct a series of experiments analyzing the system for performance and correctness of events detection.

This thesis has been concerned with a few problems in systems driven at the scale of particles. The problems dealt with here can be extended and elaborated upon in a variety of ways. In 2 we examine the dynamics of a fluid membrane in contact with a fluid containing active particles. In particular, we show that such a membrane generically enters a statistical steady state with wave-like dispersion. While the numerical results are satisfying, a one-step coarse-graining calculation, in line with [66,93], will, we expect, yield a pair of coupled stochastic differential equations (probably KPZ like at least in one dimension) with wave-like dispersion. This calculation in of interest from a theoretical point-of-view. Further, the numerical exploration of the full set of equations is also left for future work, but can be relevant to many biological systems. In 3 we show that an active fluid confined in an annular channel starts to rotate spontaneously. Further, we predict the existence of banded concentration profile. Such profiles have not yet been observed in experiments. Further, it will be interesting to study what happens to our conclusions if we include the effect of treadmilling in our calculation. In 4 we describe a solid driven by active particles. Specifically, we only concern ourselves with the polar elastomeric phase of the material. However, the questions regarding the transition into that phase are interesting and have not been explored. How exactly does a polarisation transition happen in an active polar elastomer? Is it the same as in an active nematic elastomer? What is the nature of the gelation transition in an active polar fluid? What is the dynamics of nematic defects in an elastomer? Can the presence of the elastomer prevent defect separation? We are at present trying to answer these questions. In 5 we examine the dynamics of an active fluid confined in a channel. It will be interesting to test the prediction about fluctuations in a confined active system, which we show will be normal, in experiments on highly confined actomyosin systems. In 6 we write down the coupled equations of a conformation tensor and the apolar order parameter. This is a generic framework for studying viscoelastic active fluids. A fuller study of the effect of increasing the cross-linker density in such system remains to be done, both theoretically and experimentally. In general, we have shown in the thesis that the understanding of active systems can provide a mechanistic explanation of various biological observations. However, at times the comparison between theory and biological experiments become complicated due to the inherently complicated nature of the experimental systems. Thus, for a more rigorous experimental test of the theory, it is necessary to construct cleaner reconstituted systems with possibly as few as three components. Efforts in this direction have recently borne fruit [129]. However, a complete theoretical understanding of the rich behaviour evinced in these systems is as yet lacking. We expect that the conformation tensor theory we developed in chapter 6 will provide an explanation for the anomalous rheological behaviour observed in these systems. Even in the theoretical front, lot of questions remain to be answered. The dry polar active system, described by the Toner-Tu equations have been shown to undergo a transition to a state with LRO. However, though mean-field theory predicts a second order transition [151, 152, 156], detailed numerical analysis suggests that it is actually first-order with pre-transitional solitonic bands. This has been recently examined by Chate et al. [26] who mapped it to a dynamical system, but a complete theory is still lacking. Apolar systems present another set of challenges. First, the concentration coupling with the order parameter should create similar pre-transitional effects at the order-disorder transition for this system also. This has been studied to a certain extent [133]. However, the more interesting question concerns the role of defects in apolar systems and whether they allow for the possibility of even QLRO in two dimensions. The +1/2 nematic defect has a polarity, and can thus move balistically [51, 108, 115, 149] in a dry system. However, the −1/2 defect has a three-fold symmetry [27] and its motion is thus purely diffusive. Now consider a pair of +1/2 and −1/2 defect pair that can form due to noise in the system (since it does not violate charge conservation). Depending on the configuration and the kind of activity, this defect pair can unbind at zero temperature. Unbound defects would imply that the order is short-ranged. However, it appears from detailed simulations of an agent based Vicsek-like model of active nematics, that there exists a QLRO nematic in two dimensions [111]! How does an active nematic escape being destroyed by defect unbinding? Does concentration have a major role to play? If so, does making the concentration a non-conserved, and thus fast, variable by, for example, including evaporation-deposition rules in the model studied by Chate et al. [28] destroy the QLRO? Also, does the hydrodynamic theory for Malthusian (i.e. one in which the concentration relaxes fast to a steady value) nematics show only short-ranged order, while the one in which mass is conserved show QLRO? These questions are being studied at present by simulating both the agent-based model due to Chate with evaporation-deposition and the dynamical equation for the active nematic order-parameter. These studies should clarify the role of concentration in assisting apolar order. It must be borne in mind, however, that numerical simulations of active models are more difficult than their passive counterparts due to the larger number of parameters present in the problem. In passive systems Onsager symmetry relations constrain some parameters. However, the absence of an equivalent rule for systems far away from equilibrium implies that the spatial symmetry allowed couplings will all have independent kinetic coefficients. This increases the size of the parameter space in many problems. Also, many techniques like Monte Carlo have to be carefully modified to suit such systems. A new and exciting area of research from the point of view of statistical mechanics of active systems is an examination of collective behaviour of run-and-tumble particles pioneered by Tailleur and Cates [25]. This has led to fruitful active generalisations of models of dynamic critical phenomena like model B and model H. Also, it has led to an exploration of rules for selecting a state in a region of phase coexistence – an out of equilibrium generalisation of the Maxwell construction. Another interesting avenue is building up active matter equations from microscopics. This has been done for Vicsek model by Thomas Ihle [64,65], for a simple generalisation of Vicsek-type model for both polar and apolar alignment interactions by Bertin et al. and Chate et al. [15, 16, 107], and for a model of hard rods by Marchetti et al. [10, 11]. The issues of closure still remain to be fully resolved however in deriving the macroscopic equations. A particularly exciting new system that has been recently studied extensively is a collection of chemotactic Janus particles [127]. The far-field interaction in this case does not promote polar order but state with proliferation of asters. The coarse-grained hydrodynamic equations have been derived in this case starting from a microscopic picture of colloids coated axisymetrically with a catalyst in an inhomogeneous concentration of reactants by Saha et al. [127]. Another theoretical issue that plagues the derivation of hydrodynamic equations is that of noise. So far most theories have modelled the noise as Gaussian and white, akin to equilibrium systems, but with unknown strength. However, it is likely that the noise also depends on activity, thus requiring a microscopic picture treating the active forces as stochastic quantities. It is known that multiplicative character of the noise induces interesting features at least in the case of active nematics [104]. Thus, a lot of questions need to be answered if theories of active matter have to graduate from merely offering qualitative explanations of biological experiments to becoming the prototypical theory of systems in which energy input and dissipation both occur at a scale smaller than the coarse-graining volume.

Data Stream mining is an important emerging topic in the data mining and machine learning domain. In a Data Stream setting, the data arrive continuously and often at a fast pace. Examples include credit cards transaction records, surveillances video streams, network event logs, and telecommunication records. Such types of data bring new challenges to the data mining research community. Specifically, a number of researchers have developed techniques in order to build accurate classification models against such Data Streams. Ensemble Learning, where a number of so-called base classifiers are combined in order to build a model, has shown some promise. However, a number of challenges remain. Often, the class labels of the arriving data are incorrect or missing. Furthermore, Data Stream algorithms may benefit from an online learning paradigm, where a small amount of newly arriving data is used to learn incrementally. To this end, the use of Active Learning, where the user is in the loop, has been proposed as a way to extend Ensemble Learning. Here, the hypothesis is that Active Learning would increase the performance, in terms of accuracy, ensemble size, and the time it takes to build the model. This thesis tests the validity of this hypothesis. Namely, we explore whether augmenting Ensemble Learning with an Active Learning component benefits the Data Stream Learning process. Our analysis indicates that this hypothesis does not necessarily hold for the datasets under consideration. That is, the accuracies of Active Ensemble Learning are not statistically significantly higher than when using normal Ensemble Learning. Rather, Active Learning may even cause an increase in error rate. Further, Active Ensemble Learning actually results in an increase in the time taken to build the model. However, our results indicate that Active Ensemble Learning builds accurate models against much smaller ensemble sizes, when compared to the traditional Ensemble Learning algorithms. Further, the models we build are constructed against small and incrementally growing training sets, which may be very beneficial in a real time Data Stream setting.

Le fonctionnement des organismes vivants requiert la production deforces à grande échelle, pour des processus biologiques aussi diversque la motilité cellulaire, le développement embryonnaire, lacicatrisation ou encore la contraction musculaire. Dans de telssystèmes, les forces générées à l'échelle moléculaire par des moteursprotéiques sont transmises par des réseaux de fibres désordonnés,menant à des tensions actives à grande échelle. Les propriétésmacroscopiques passives de ces réseaux de fibres sont biencaractérisées. En revanche, ce problème de production de stress pardes unités actives microscopiques n'est pas résolu. Cette Thèseprésente une étude approfondie, par des méthodes théoriques etnumériques, de la transmission de forces dans les réseaux élastiquesde biopolymères. Je montre que la réponse linéaire, à faible force,des réseaux est remarquablement simple : elle est déterminée par laseule la géométrie des unités actives exerçant les forces. Aucontraire, lorsque les forces actives sont suffisamment importantespour provoquer le flambage non-linéaire des fibres, ces forces sontrectifiées par le réseau, et deviennent isotropiquementcontractiles. La contraction émergente qui en résulte est amplifiéepar la transmission de forces non-linéaire à travers le réseau. Cetteamplification du stress macroscopique est renforcée par le caractèredésordonnée du réseau, mais sature lorsque la densité d'unités activesest grande. Nos prédictions sont en accord quantitatifs avec desrésultats expérimentaux sur des tissus reconstitués et des réseauxd'actomyosine in vitro, et apportent un éclairage nouveau surl'influence de l'architecture microscopique des réseaux sur structuredes stress à l'échelle de la cellule et du tissu. / Large-scale force generation is essential for biological functionssuch as cell motility, embryonic development, wound healing and musclecontraction. In these processes, forces generated at the molecularlevel by motor proteins are transmitted by disordered fiber networks,resulting in large-scale active stresses. While fiber networks arewell characterized macroscopically, this stress generation bymicroscopic active units is not well understood. In this Thesis, Ipresent a comprehensive theoretical and numerical study of forcetransmission in elastic fiber networks. I show that the linear,small-force response of the networks is remarkably simple, as themacroscopic active stress depends only on the geometry of theforce-exerting unit. In contrast, as non-linear buckling occurs aroundthese units, local active forces are rectified towards isotropiccontraction, making the local geometry of force exertion irrelevant.This emergent contractility is amplified by non-linear forcetransmission through the network. This stress amplification isreinforced by the networks' disordered nature, but saturates for highdensities of active units. Our predictions are quantitativelyconsistent with experiments on reconstituted tissues and actomyosinnetworks, and that they shed light on the role of the networkmicrostructure in shaping active stresses in cells and tissue.