The phenomenal growth in the use of the World Wide Web often places a heavy load on networks and servers, threatening to increase Web server response time and raising scalability issues for both the network and the server. With the advances in the field of optical networking and the increasing use of broadband technologies like cable modems and DSL, the server and not the network, is more likely to be the bottleneck. Many clients are willing to receive a degraded, less resource intensive version of the requested content as an alternative to connection failures. In this thesis, we present an adaptive content delivery system that transparently switches content depending on the load on the server in order to serve more clients. Our system is designed to work for dynamic Web pages and streaming multimedia traffic, which are not currently supported by other adaptive content approaches. We have designed a system which is capable of quantifying the load on the server and then performing the necessary adaptation. We designed a streaming MPEG server and client which can react to the server load by scaling the quality of frames transmitted. The main benefits of our approach include: transparent content switching for content adaptation, alleviating server load by a graceful degradation of server performance and no requirement of modification to existing server software, browsers or the HTTP protocol. We experimentally evaluate our adaptive server system and compare it with an unadaptive server. We find that adaptive content delivery can support as much as 25% more static requests, 15% more dynamic requests and twice as many multimedia requests as a non-adaptive server. Our, client-side experiments performed on the Internet show that the response time savings from our system are quite significant.

Heavy-tailed workload distributions are commonly experienced in many areas of distributed computing. Such workloads are highly variable, where a small number of very large tasks make up a large proportion of the workload, making the load very hard to distribute effectively. Traditional task assignment policies are ineffective under these conditions as they were formulated based on the assumption of an exponentially distributed workload. Size-based task assignment policies have been proposed to handle heavy-tailed workloads, but their applications are limited by their static nature and assumption of prior knowledge of a task's service requirement. This thesis analyses existing approaches to load distribution under heavy-tailed workloads, and presents a new generalised task assignment policy that significantly improves performance for many distributed applications, by intelligently addressing the negative effects on performance that highly variable workloads cause. Many problems associated with the modelling and optimisations of systems under highly variable workloads were then addressed by a novel technique that approximated these workloads with simpler mathematical representations, without losing any of their pertinent original properties. Finally, we obtain advance queuing metrics (such as the variance of key measurements like waiting time and slowdown that are difficult to obtain analytically) through rigorous simulation.

We consider the problem of estimating the reliability of bundles constructed of several fibres, given a particular kind of censored data. The bundles consist of several fibres which have their own independent identically dis-tributed failure stresses (i.e.the forces that destroy the fibres). The force applied to a bundle is distributed between the fibres in the bundle, accord-ing to a load-sharing model. A bundle with these properties is an example of a load-sharing system. Ropes constructed of twisted threads, compos-ite materials constructed of parallel carbon fibres, and suspension cables constructed of steel wires are all examples of load-sharing systems. In par-ticular, we consider bundles where load-sharing is described by either the Equal load-sharing model or the more general Local load-sharing model. In order to estimate the cumulative distribution function of failure stresses of bundles, we need some observed data. This data is obtained either by testing bundles or by testing individual fibres. In this thesis, we develop several theoretical testing methods for both fibres and bundles, and related methods of statistical inference. Non-parametric and parametric estimators of the cumulative distribu-tion functions of failure stresses of fibres and bundles are obtained from different kinds of observed data. It is proved that most of these estimators are consistent, and that some are strongly consistent estimators. We show that resampling, in this case random sampling with replacement from sta-tistically independent portions of data, can be used to assess the accuracy of these estimators. Several numerical examples illustrate the behavior of the obtained estimators. These examples suggest that the obtained estimators usually perform well when the number of observations is moderate.

Abstract Bleaching is an essential process in chemical pulp production for better pulp brightness and longer life expectancy. However, it causes costs such as chemicals, energy, equipment, and loss of yield. Non-linear reactions and several process variables, with interactions, make large plants complicated to model and optimize. As an expensive process bleaching has been a natural target of optimization, but there is still the need to either improve these methods or consider the optimization problem from a new point of view. The aim of this thesis was to develop production optimization methods for pulp bleaching, so that they are practical, usable on-line, easy to tune, and transferable. According to our assumption, neural networks could provide a practical optimization method by combining analytical knowledge with real data. In this kind of problem, the load sharing concept, recognizing interactions in chemical usage and the serial multi-stage nature of the process can simplify the task. The related work in bleaching optimization was studied as well as multi-stage serial process solving in principle, related optimization methods and especially neural networks in optimization. The data were collected during normal mill operation and modeled using neural networks. Optimization was performed based on visualizing the neural network models. The results showed that backpropagation neural networks are capable of modeling parts of the bleach plant and also the entire bleaching operation to such an extent that they are useful in the optimization. The modeling and the tuning can be performed without a profound knowledge of the system, but the process is slower and less reliable. Moving a trained neural network to another mill is inadvisable. It is more reasonable just to transfer the knowledge of variables and network structure. The important factor in on-line production optimization is the stabilization of the disturbances and a well-controlled operation towards a more economical state. Generally, more than half of the total chemicals should be used in the first bleaching stage D0 and the remaining load should be divided so that the dosage at the D1 is about 30% higher than in the D2 stage.

In near future, robots are envisioned to work alongside humans in professional anddomestic environments without significant restructuring of workspace. Roboticsystems in such setups must be adept at observation, analysis and rational de-cision making. To coexist in an environment, humans and robots will need tointeract and cooperate for multiple tasks. A fundamental such task is the manip-ulation of large objects in work environments which requires cooperation betweenmultiple manipulating agents for load sharing. Collaborative manipulation hasbeen studied in the literature with the focus on multi-agent planning and controlstrategies. However, for a collaborative manipulation task, grasp planning alsoplays a pivotal role in cooperation and task completion.In this work, a novel approach is proposed for collaborative grasping and manipu-lation of large unknown objects. The manipulation task was defined as a sequenceof poses and expected external wrench acting on the target object. In a two-agentmanipulation task, the proposed approach selects a grasp for the second agentafter observing the grasp location of the first agent. The solution is computed ina way that it minimizes the grasp wrenches by load sharing between both agents.To verify the proposed methodology, an online system for human-robot manipu-lation of unknown objects was developed. The system utilized depth informationfrom a fixed Kinect sensor for perception and decision making for a human-robotcollaborative lift-up. Experiments with multiple objects substantiated that theproposed method results in an optimal load sharing despite limited informationand partial observability.

Wind farm construction is increasing progressively, to cope-up with the current global energy scenario. The advantage of clean energy and sustainability helps wind turbine construction to flourish rapidly. Location of wind turbines is independent of foundation soil condition but depends on the wind speeds and socio-environment issues. Hence, a construction sites may not be favorable in terms of geotechnical demands. The taller wind towers facilitate the generation of high energy production, which will increase loads on the foundation, and eventually increase the dimension of the foundation. Hence, the choice of a suitable foundation system is necessary for geotechnical engineer to design tall wind towers. This study aims to analyze different foundation types e.g., raft/mat foundation, pile group foundation, and piled raft/mat foundation using analytical calculation verified with numerical models using PLAXIS 3D software. The foundation for steel wind turbine towers 100 m high was designed for different types of soils e.g., soft clayey soil, medium-stiff clayey soil, stiff clayey soil, and sandy soil. The design wind speed was taken from the ASCE 7-10 (2010) standard for Occupancy Category III and IV Buildings and Other Structures, as the Illinois region falls in that category. The parametric study was performed by varying the diameter of raft/mat, wind speed, number of piles, and soil types to evaluate the settlement in any type of foundation with load sharing proportion in piled raft/mat foundation. First, the raft/mat foundation design was carried out manually by changing the diameter of 15 m, 20 m, 25 m, 30 m, and 35 m, and changing load by considering different wind speed. Then the foundation was modeled using PLAXIS 3D software with a raft/mat diameter of 25 m, 30 m, and 35 m only, by considering the eccentricity and factor of safety criteria. With the increase in wind speed, the differential settlement on the raft/mat foundation was found to be increased. However, the increase in diameter of raft/mat caused the reduction in differential settlement. Soft clayey soil was found to be more sensitive than other soils used in the present study. For the same diameter of raft/mat, applied the same wind load, the differential settlement of foundation in soft clayey soil was found to be 6-10 times higher than the sandy soil.The position of piles was fixed based on the spacing criteria in the pile group foundation. The number of piles used in this study were 23, 32, and 46. Settlement was found to be varied with the number of piles in all soils used in this study. The lateral deflection for soft clayey soil decreased to half, when number of piles increased from 23 to 46. The differential settlement was found to be increased with the increase in wind speed in pile group foundation. Raft/mat foundation settlement was found to be 4 to 6 times higher than the settlement in pile group foundation in any soils, used in this study, for a given wind speed.The result of piled raft/mat foundation showed that the majority of the total load is shared by the piles (i.e., 60% to 94%) and remaining load is shared by the raft/mat (i.e., 6% to 40%), based on the stiffness of raft/mat and piles as well as pile-soil-pile interaction. The increase in wind speed in the wind turbines increased the differential settlement of piled raft/mat foundation in all soils. Similarly, the lateral deflection also increased with the increase in wind speed in pile raft/mat foundation in all soils. The PLAXIS 3D analysis revealed that the differential settlement in soft clayey soil was 1.5 to 2.0 times higher than the settlement in sandy soil.The validation of numerical modeling was carried out by the raft/mat foundation using Boussinesq’s theory and calculating settlement for single pile and group pile foundation. The current study showed that the soft clayey soil and medium-stiff clayey soil favor deep foundation, like pile group and piled raft/mat rather than shallow foundation, like raft/mat foundation. The results obtained from both analytical calculation and numerical modeling was found to be approximately matching. This study will help local construction company and geotechnical engineer to guide a proper foundation design of tall onshore wind turbine.

A simulation model which incorporates the statistical- and numerical-based Lattice Green Function Local Load Sharing Model and a Fracture Mechanics-based Residual Strength Model has been developed. The model simulates creep rupture by imposing a fixed load of constant stress on the composite over the simulation duration. Simulation of the fatigue of glass fiber-reinforced composites is achieved by replacing the constant stress parameter in the model with a sinusoidal wave function. Results from the creep rupture model using fused silica fiber parameters, compare well with S-2 glass/epoxy systems. Results using Mandell's postulate that fatigue failure in glass fiber-reinforced polymeric composites is a fiber-dominated mechanism, with a characteristic slope of 10 %UTS/decade are consistent with available experimental data. The slopes of fatigue curves for simulated composites for three frequencies namely: 2, 5 and 10 Hz are within 12-14 %UTS/decade compared with that of 10.6-13.0%UTS/decade for unidirectionl glass reinforced composites (epoxy and vinyl ester) obtained from Demers' [40] data. / Master of Science

DC power distribution has gained popularity in sustainable buildings, renewable energy utilization, transportation electrification and high-efficiency data centers. This dissertation focuses on two aspects of facilitating the application of dc systems: (a) system-level control to improve load sharing, voltage regulation and efficiency; (b) design of a high-efficiency interface converter to connect dc microgrids with the existing low-voltage ac distributions, with a special focus on common-mode (CM) voltage attenuation. Droop control has been used in dc microgrids to share loads among multiple sources. However, line resistance and sensor discrepancy deteriorate the performance. The quantitative relation between the droop voltage range and the load sharing accuracy is derived to help create droop design guidelines. DC system designers can use the guidelines to choose the minimum droop voltage range and guarantee that the sharing error is within a defined range even under the worst cases. A nonlinear droop method is proposed to improve the performance of droop control. The droop resistance is a function of the output current and increases when the output current increases. Experiments demonstrate that the nonlinear droop achieves better load sharing under heavy load and tighter bus voltage regulation. The control needs only local information, so the advantages of droop control are preserved. The output impedances of the droop-controlled power converters are also modeled and measured for the system stability analysis. Communication-based control is developed to further improve the performance of dc microgrids. A generic dc microgrid is modeled and the static power flow is solved. A secondary control system is presented to achieve the benefits of restored bus voltage, enhanced load sharing and high system efficiency. The considered method only needs the information from its adjacent node; hence system expendability is guaranteed. A high-efficiency two-stage single-phase ac-dc converter is designed to connect a 380 V bipolar dc microgrid with a 240 V split-phase single-phase ac system. The converter efficiencies using different two-level and three-level topologies with state-of-the-art semiconductor devices are compared, based on which a two-level interleaved topology using silicon carbide (SiC) MOSFETs is chosen. The volt-second applied on each inductive component is analyzed and the interleaving angles are optimized. A 10 kW converter prototype is built and achieves an efficiency higher than 97% for the first time. An active CM duty cycle injection method is proposed to control the dc and low-frequency CM voltage for grounded systems interconnected with power converters. Experiments with resistive and constant power loads in rectification and regeneration modes validate the performance and stability of the control method. The dc bus voltages are rendered symmetric with respect to ground, and the leakage current is reduced. The control method is generalized to three-phase ac-dc converters for larger power systems. / Ph. D.

This thesis focuses on the development of a low power energy harvesting system specifically targeted for wireless sensor nodes (WSN) and wireless body area network (WBAN) applications. The idea for the system is derived from the operation of a micro-grid and therefore is termed as a pico-grid and it is capable of simultaneously delivering power from multiple and multitype energy harvesters to the load at the same time, through the proposed parallel load sharing mechanism achieved by a voltage droop control method. Solar panels and thermoelectric generator (TEG) are demonstrated as the main energy harvesters for the system. Since the magnitude of the output power of the harvesters is time-varying, the droop gain in the droop feedback circuitry should be designed to be dynamic and self-adjusted according to this variation. This ensures that the maximum power is capable to be delivered to the load at all times. To achieve this, the droop gain is integrated with a light dependent resistor (LDR) and thermistor whose resistance varies with the magnitude of the source of energy for the solar panel and TEG, respectively. The experimental results demonstrate a successful variation droop mechanism and all connected sources are able to share equal load demands between them, with a maximum load sharing error of 5 %. The same mechanism is also demonstrated to work for maximum power point tracking (MPPT) functionality. This concept can potentially be extended to any other types of energy harvester. The integration of energy storage elements becomes a necessity in the pico-grid, in order to support the intermittent and sporadic nature of the output power for the harvesters. A rechargeable battery and supercapacitor are integrated in the system, and each is accurately designed to be charged when the loading in the system is low and discharged when the loading in the system is high. The dc bus voltage which indicates the magnitude of the loading in the system is utilised as the signal for the desired mode of operation. The constructed system demonstrates a successful operation of charging and discharging at specific levels of loading in the system. The system is then integrated and the first wearable prototype of the pico-grid is built and tested. A successful operation of the prototype is demonstrated and the load demand is shared equally between the source converters and energy storage. Furthermore, the pico-grid is shown to possess an inherent plug-and-play capability for the source and load converters. Few recommendations are presented in order to further improve the feasibility and reliability of the prototype for real world applications. Next, due to the opportunity of working with a new semiconductor compound and accessibility to the fabrication facilities, a ZnON thin film diode is fabricated and intended to be implemented as a flexible rectifier circuit. The fabrication process can be done at low temperature, hence opening up the possibility of depositing the device on a flexible substrate. From the temperature dependent I-V measurements, a novel method of extracting important parameters such as ideality factor, barrier height, and series resistance of the diode based on a curve fitting method is proposed. It is determined that the ideality factor of the fabricated diode is high (> 2 at RT), due to the existence of other transport mechanism apart from thermionic emission that dominates the conduction process at lower temperature. It is concluded that the high series resistance of the fabricated diode (3.8 kΩ at RT) would mainly hinder the performance of the diode in a rectifier circuit.

Researches on biological adhesive systems in nature have changed a perspective view on adhesion that it is not only the area of surface chemistry, but also mechanics of interfacial geometry which can significantly effect on fracture strength and load distribution on the contact interface. Various synthetic fibrillar adhesives in previous works have shown enhanced interfacial bond strength with the capacity of adhesion control by exploiting mechanical deformation of the elastomeric fibrillar structures inspired by geckos. However, control of the interfacial load distribution has been focused on the size of micro-contact with single or a few of micro-/nano-fibers on planar surface, and not for a large contact area on complex three-dimensional (3D) surfaces. This thesis work aims at investigating principles of the interfacial load distribution control in multi-scale, ranging from micro-contact with single micro-fiber to a centimeter-scale contact with a membrane-backed micro-fiber array on non-planar 3D surfaces. The findings are also applied for developing a soft robotic gripper capable of grasping a wide range of complex objects in size, shape, and number, expanding the area of practical applications for bio-inspired adhesives in transfer printing, robotic manipulators, and mobile robots. This paper comprises three main works. First, we investigate the effect of tip-shapes on the interfacial load sharing of mushroom-shaped micro-fibrillar adhesives with precisely defined tipgeometries using high resolution 3D nano-fabrication technique. For a large area of non-planar contact interface, we fabricate fibrillar adhesives on a membrane (FAM) by integrating micro-fibers with a soft backing, which enables robust and controllable adhesion on 3D surfaces. Picking and releasing mechanism for the maximal controllability in adhesion are discussed. Finally, we propose a soft robotic architecture which can control the interfacial load distribution for the FAM on 3D surfaces, solving an inherit dilemma between conformability and high fracture strength with the equal load sharing on complex non-planar 3D surfaces.