Facing the pressure of population growth and global warming, this dissertation provided an array of innovative carbon-efficient wastewater treatment technologies for resource recovery, process intensification, and anammox featured next generation biological nutrient removal (BNR) technologies. These technologies aim to supplant traditional carbon-intensive treatment processes with more sustainable alternatives. To this end, the dissertation first comprehensively reviewed what resources can be recovered from wastewater, and how these valuable resources can contribute to the carbon neutrality in water resource reclamation facilities (WRRFs) and help achieve sustainable society development. Then, the effect of mixed liquor recycle (MLR) configurations on the process intensification through continuous-flow aerobic granulation was explored in plug flow reactors. The results demonstrated that MLR configuration could hinder the sludge granulation, but the hindrance could be alleviated to some extent by its location change. In order to eliminate the energy consuming MLR, endogenous denitrification was taken advantage through a synergistic integration with partial nitrification, partial denitrification anammox (PdNA), and enhanced biological phosphorus removal (EBPR). This idea was tested in a pilot setup treating real primary effluent under highly variable influent conditions and low temperatures. The results showcased substantial carbon savings while meeting the stringent effluent requirements. To take a deeper dive into the PdNA performance and the underlying mechanisms, two parallel pilot-scale moving bed biofilm reactor (MBBR) treatment trains fed with methanol and glycerol, respectively, were operated in a local WRRF. Their efficacies in achieving stringent nutrient removal targets and carbon savings were compared. The impacts of operational conditions on the mechanisms and performance were elucidated. In the culmination of this dissertation, a sidestream process intensification and resource recovery technique, namely thermal hydrolysis pretreatment (THP) enhanced anaerobic digestion (AD), was experimented to compare the efficiencies between thermophilic and mesophilic AD when integrated with THP. To sum up, this dissertation not only advanced our understanding of carbon-efficient wastewater treatment processes but also laid the groundwork for their practical implementation, contributing to the global effort towards sustainability. / Doctor of Philosophy / Wastewater treatment consumes 3-4% of the energy produced in the U.S. and contributes to approximately 1.6% global greenhouse gas emissions. This dissertation aims to advance a series of carbon-efficient technologies specifically tailored for sustainable wastewater treatment. To this end, a variety of valuable resources that can be recovered or reused in wastewater treatment plants was firstly reviewed. Then, an advanced technology that can turn dispersed bacteria into bacteria aggregates was tested with real wastewater in a local wastewater treatment plant. Although these bacteria aggregates allow more wastewater to be treated with less small footprint, which was great, it was realized from this study that the formation of these bacteria aggregates was hindered by the nitrate water recycle which has been commonly practiced for using influent carbon for nitrogen removal. This nitrate water recycle consumed excessive energy for its high flow rate. To save this energy, a novel bioprocessing design was developed to eliminate the need for this nitrate water recycle by using carbon stored in bacterial cells. This new design also incorporated phosphorus recovery capacity and a low carbon nitrogen removal technique into one consolidated system to create an all-in-one solution to meet the stringent wastewater treatment requirement. This low carbon nitrogen removal technique harnessed a special group of bacteria that can use ammonia to reduce nitrite to nitrogen gas. Hence, only minor carbon source needs to be provided to reduce nitrate to nitrite for these bacteria to utilize. Two types of carbon sources, namely methanol and glycerol, were compared in a pilot-scale study to understand their efficiencies in generating nitrite. Results indicated that although both types of carbon sources can work, methanol is better suited for low strength wastewater treatment. These results provided an engineering basis for the full-scale application of the technology in the same wastewater treatment plant where the pilot study was performed. Besides liquid treatment, a carbon efficient solid treatment technology was also studied. The bottleneck constraining the rate of sewage sludge conversion to flammable menthane gas was identified, which provided engineering guidance for the design of the solid treatment process that can destroy more sewage sludge within smaller reactor spaces. In essence, this dissertation offers promising solutions for modern wastewater treatment plants to achieve low carbon wastewater treatment without compromising the treatment performance.

Recent advancements in deep learning have demonstrated its incredible ability to learn generalizable patterns and relationships automatically from data in a number of mainstream applications. However, the generalization power of deep learning methods largely comes at the costs of working with very large datasets and using highly compute-intensive models. Many applications cannot afford these costs needed to ensure generalizability of deep learning models. For instance, obtaining labeled data can be costly in scientific applications, and using large models may not be feasible in resource-constrained environments involving portable devices. This dissertation aims to improve efficiency in machine learning by exploring different ways to learn generalizable neural networks that require less labeled data and computational resources. We demonstrate that using physics supervision in scientific problems can reduce the need for labeled data, thereby improving data efficiency without compromising model generalizability. Additionally, we investigate the potential of transfer learning powered by transformers in scientific applications as a promising direction for further improving data efficiency. On the computational efficiency side, we present two efforts for increasing parameter efficiency of neural networks through novel architectures and structured network pruning. / Doctor of Philosophy / Deep learning is a powerful technique that can help us solve complex problems, but it often requires a lot of data and resources. This research aims to make deep learning more efficient, so it can be applied in more situations. We propose ways to make the deep learning models require less data and less computer power. For example, we leverage the physics rules as additional information for training the neural network to learn from less labeled data and we use a technique called transfer learning to leverage knowledge from data that is from other distribution. Transfer learning may allow us to further reduce the need for labeled data in scientific applications. We also look at ways to make the deep learning models use less computational resources, by effectively reducing their sizes via novel architectures or pruning out redundant structures.

In recent years, significant progress has been made in improving the performance of unmanned air vehicles in terms of aerodynamic performance, endurance, autonomy, and the capability of on-board sensor packages. UAVs are now a vital part of both military actions and scientific research efforts. One of the newest classes of UAV is the high altitude long endurance or HALE UAV. This thesis considers the high-level control problem for a unique HALE mission involving cooperative solar powered UAVs. Specifically addressed is energy efficient path planning for vehicles that physically link together in flight to form a larger, more energy efficient HALE vehicle. Energy efficient docking is developed for the case of multiple vehicles at high altitude with negligible wind. The analysis considers a vehicle governed by a kinematic motion model with bounded turn rate in planar constant altitude flight. Docking is demonstrated using a platform-in-the-loop simulator which was developed to allow virtual networked vehicles to perform decentralized path planning and estimation of all vehicle states. Vehicle behavior is governed by a status which is commanded by a master computer and communication between vehicles is intermittent depending on each vehicle's assessment of situational awareness. Docking results in a larger vehicle that consumes energy at 21% of the rate of an individual vehicle and increases vehicle range by a factor of three without considering solar recharging. / Master of Science

The production of polyethylene terephthalate (PET) textiles, commonly known as polyester, is an energy and water-intensive process, particularly during the dyeing and finishing stages, leading to significant carbon emissions and wastewater filled with toxic chemicals. Conventional dyeing processes consume large volumes of water and energy, making them environmentally harmful. Innovative methods like hydraulic spray atomizer offer a more sustainable alternative by reducing water and chemical usage, thus minimizing waste and environmental impact. This study explores the hydraulic spray dyeing process parameters, including pre-treatment, dyeing, pre-drying, and fixation steps for three different polyester fabrics. The objective is to achieve resource-efficient dyeing with comparable results to traditional methods, and to compare effectiveness of a combined pre-treatment and dyeing approach with a two-step spray application of pre-treatment and dyeing. Pre-treatment of the polyester fabrics with chemical hydrophilizing agents by spray application showed to improve the hydrophilic character and wetting capacity of three polyester fabrics—P75, P600, and Kibo based on reduced water contact angle measurements and increased vertical wicking rates. A pre-treatment also showed to enhance the K/S values of the three different polyester fabrics. The addition of a pre-drying step showed some indication of reducing disperse dye migration, and enhanced color strength of the Kibo fabrics. Spray dyed samples maintained dyeing quality comparable to padded samples. One-step spray processes demonstrated comparable or improved color properties and durability to fastness to washing and abrasion in comparison to two-step processes. The results demonstrate that the hydraulic spray atomizing system is viable for both dyeing and pre-treatment of polyester fabrics. Furthermore, this lays the groundwork for innovation in wet textile processes of polyester fabrics using this resource-efficient alternative, aiming towards sustainable textile production and dyeing.

Investors have developed and used a range of investment strategies to generate a higherreturn than the overall market. Among these strategies, value and growth investing aretwo strategies that have become especially popular within the investment community.The difference between the two strategies originates from their differing perspectives onvalue ratios, where growth investors search for stocks with higher ratios on metrics likeprice-to-earnings (P/E) and price-to-book (P/B), called growth stocks, while valueinvestors seek stocks with lower ratios, called value stocks. The main purpose of thisstudy is to determine whether value or growth stocks provide a superior risk-adjustedreturn to offer investors an updated insight on portfolio allocation. The secondary purposeis to capture how resilient or sensitive the two types of stocks are to market volatility, toidentify characteristics that make certain compositions of stocks more effective duringdifferent periods. The sample consists of firms included in the S&P 500 index and thestocks are classified into value or growth stocks using the P/E ratio and the P/B ratio.Tests are performed each year between 2012 and 2023 to see how they perform, and withthe Sharpe ratio we are able to compare the two stock types based on their risk-adjustedreturn. Early research on value and growth investing came to the same conclusion, that valuestocks give a higher return than growth stocks, which has been the general view on thetwo strategies. More recent studies have identified a potential shift in the previous view,with indications that growth stocks perform better, and in recent years, firms in the techoriented business have seen high ratios, but at the same time they have generated highreturns. The empirical results show that during the time period studied, growth stocksoutperform value stocks in some years, value stocks outperform growth stocks in others,and in some, no statistical difference between the two is found. Over the whole period,from 2012 to 2023, we find that growth stocks have provided a higher risk-adjusted returncompared to value stocks, aligning with the most recent studies and challenging theprevious view that value stocks perform better. During volatile times, in this studyidentified as 2020 to 2022 during the Covid-19 crisis, the empirical result show that involatile market conditions, value stock perform better and is the better alternative for riskadjusted return.

[EN] Bridges, as an important component of infrastructure, are expected to meet all the requirements for a modern society. Traditionally, the primary aim in bridge design has been to achieve the lowest cost while guaranteeing the structural efficiency. However, concerns regarding building a more sustainable future have change the priorities of society. Ecological and durable structures are increasingly demanded. Under these premises, heuristic optimization methods provide an effective alternative to structural designs based on experience. The emergence of new materials, structural designs and sustainable criteria motivate the need to create a methodology for the automatic and accurate design of a real post-tensioned concrete bridge that considers all these aspects. For the first time, this thesis studies the efficient design of post-tensioned concrete box-girder road bridges from a sustainable point of view. This research integrates environmental, safety and durability criteria into the optimum design of the bridge. The methodology proposed provides multiple trade-off solutions that hardly increase the cost and achieve improved safety and durability. Likewise, this approach quantifies the sustainable criteria in economic terms, and evaluates the effect of these criteria on the best values of the variables. In this context, a multi-objective optimization is formulated to provide multiple trade-off and high-performing solutions that balance economic, ecologic and societal goals. An optimization design program selects the best geometry, concrete type, reinforcement and post-tensioning steel that meet the objectives selected. A three-span continuous box-girder road bridge located in a coastal region is selected for a case study. This approach provides vital knowledge about this type of bridge in the sustainable context. The life-cycle perspective has been included through a lifetime performance evaluation that models the bridge deterioration process due to chloride-induced corrosion. The economic, environmental and societal impacts of maintenance actions required to extend the service life are examined. Therefore, the proposed goals for an efficient design have been switch from initial stage to life-cycle consideration. Faced with the large computational time of multi-objective optimization and finite-element analysis, artificial neural networks (ANNs) are integrated in the proposed methodology. ANNs are trained to predict the structural response based on the design variables, without the need to analyze the bridge response. The multi-objective optimization problem results in a set of trade-off solutions characterized by the presence of conflicting objectives. The final selection of preferred solutions is simplified by a decision-making technique. A rational technique converts a verbal pairwise comparison between criteria with a degree of uncertainty into numerical values that guarantee the consistency of judgments. This thesis gives a guide for the sustainable design of concrete structures. The use of the proposed approach leads to designs with lower life-cycle cost and emissions compared to general design approaches. Both bridge safety and durability can be improved with a little cost increment by choosing the correct design variables. In addition, this methodology is applicable to any type of structure and material. / [ES] Los puentes, como parte importante de una infraestructura, se espera que reúnan todos los requisitos de una sociedad moderna. Tradicionalmente, el objetivo principal en el diseño de puentes ha sido lograr el menor coste mientras se garantiza la eficiencia estructural. Sin embargo, la preocupación por construir un futuro más sostenible ha provocado un cambio en las prioridades de la sociedad. Estructuras más ecológicas y duraderas son cada vez más demandadas. Bajo estas premisas, los métodos de optimización heurística proporcionan una alternativa eficaz a los diseños estructurales basados en la experiencia. La aparición de nuevos materiales, diseños estructurales y criterios sostenibles motivan la necesidad de crear una metodología para el diseño automático y preciso de un puente real de hormigón postesado que considere todos estos aspectos. Por primera vez, esta tesis estudia el diseño eficiente de puentes de hormigón postesado con sección en cajón desde un punto de vista sostenible. Esta investigación integra criterios ambientales, de seguridad estructural y durabilidad en el diseño óptimo del puente. La metodología propuesta proporciona múltiples soluciones que apenas encarecen el coste y mejoran la seguridad y durabilidad. Al mismo tiempo, se cuantifica el enfoque sostenible en términos económicos, y se evalúa el efecto que tienen dichos criterios en el valor óptimo de las variables. En este contexto, se formula una optimización multiobjetivo que proporciona soluciones eficientes y de compromiso entre los criterios económicos, ecológicos y sociales. Un programa de optimización del diseño selecciona la mejor combinación de geometría, tipo de hormigón, armadura y postesado que cumpla con los objetivos seleccionados. Se ha escogido como caso de estudio un puente continuo en cajón de tres vanos situado en la costa. Este método proporciona un mayor conocimiento sobre esta tipología de puentes desde un punto de vista sostenible. Se ha estudiado el ciclo de vida a través de la evaluación del deterioro estructural del puente debido al ataque por cloruros. Se examina el impacto económico, ambiental y social que produce el mantenimiento necesario para extender la vida útil del puente. Por lo tanto, los objetivos propuestos para un diseño eficiente han sido trasladados desde la etapa inicial hasta la consideración del ciclo de vida. Para solucionar el problema del elevado tiempo de cálculo debido a la optimización multiobjetivo y el análisis por elementos finitos, se han integrado redes neuronales en la metodología propuesta. Las redes neuronales son entrenadas para predecir la respuesta estructural a partir de las variables de diseño, sin la necesidad de analizar el puente. El problema de optimización multiobjetivo se traduce en un conjunto de soluciones de compromiso que representan objetivos contrapuestos. La selección final de las soluciones preferidas se simplifica mediante una técnica de toma de decisiones. Una técnica estructurada convierte los juicios basados en comparaciones por pares de elementos con un grado de incertidumbre en valores numéricos que garantizan la consistencia de dichos juicios. Esta tesis proporciona una guía que extiende y mejora las recomendaciones sobre el diseño de estructuras de hormigón dentro del contexto de desarrollo sostenible. El uso de la metodología propuesta lleva a diseños con menor coste y emisiones del ciclo de vida, comparado con diseños que siguen metodologías generales. Los resultados demuestran que mediante una correcta elección del valor de las variables se puede mejorar la seguridad y durabilidad del puente con un pequeño incremento del coste. Además, esta metodología es aplicable a cualquier tipo de estructura y material. / [CA] Els ponts, com a part important d'una infraestructura, s'espera que reunisquen tots els requisits d'una societat moderna. Tradicionalment, l'objectiu principal en el disseny de ponts ha sigut aconseguir el menor cost mentres es garantix l'eficiència estructural. No obstant això, la preocupació per construir un futur més sostenible ha provocat un canvi en les prioritats de la societat. Estructures més ecològiques i durables són cada vegada més demandades. Davall estes premisses, els mètodes d'optimització heurística proporcionen una alternativa eficaç als dissenys estructurals basats en l'experiència. L'aparició de nous materials, dissenys estructurals i criteris sostenibles motiven la necessitat de crear una metodologia per al disseny automàtic i precís d'un pont real de formigó posttesat que considere tots estos aspectos. Per primera vegada, esta tesi estudia el disseny eficient de ponts de formigó posttesat amb secció en calaix des d'un punt de vista sostenible. Esta investigació integra criteris ambientals, de seguretat estructural i durabilitat en el disseny òptim del pont. La metodologia proposada proporciona múltiples solucions que a penes encarixen el cost i milloren la seguretat i durabilitat. Al mateix temps, es quantifica l'enfocament sostenible en termes econòmics, i s'avalua l'efecte que tenen els dits criteris en el valor òptim de les variables. En este context, es formula una optimització multiobjetivo que proporciona solucions eficients i de compromís entre els criteris econòmics, ecològics i socials. Un programa d'optimització del disseny selecciona la millor geometria, tipus de formigó, armadura i posttesat que complisquen amb els objectius seleccionats. S'ha triat com a cas d'estudi un pont continu en calaix de tres vans situat en la costa. Este mètode proporciona un major coneixement sobre esta tipologia de ponts des d'un punt de vista sostenible. S'ha estudiat el cicle de vida a través de l'avaluació del deteriorament estructural del pont a causa de l'atac per clorurs. S'examina l'impacte econòmic, ambiental i social que produïx el manteniment necessari per a estendre la vida útil del pont. Per tant, els objectius proposats per a un disseny eficient han sigut traslladats des de l'etapa inicial fins a la consideració del cicle de vida. Per a solucionar el problema de l'elevat temps de càlcul degut a l'optimització multiobjetivo i l'anàlisi per elements finits, s'han integrat xarxes neuronals en la metodologia proposada. Les xarxes neuronals són entrenades per a predir la resposta estructural a partir de les variables de disseny, sense la necessitat d'analitzar el pont. El problema d'optimització multiobjetivo es traduïx en un conjunt de solucions de compromís que representen objectius contraposats. La selecció final de les solucions preferides se simplifica per mitjà d'una tècnica de presa de decisions. Una tècnica estructurada convertix els juís basats en comparacions per parells d'elements amb un grau d'incertesa en valors numèrics que garantixen la consistència dels dits juís. Esta tesi proporciona una guia que estén i millora les recomanacions sobre el disseny d'estructures de formigó dins del context de desenrotllament sostenible. L'ús de la metodologia proposada porta a dissenys amb menor cost i emissions del cicle de vida, comparat amb dissenys que seguixen metodologies generals. Els resultats demostren que per mitjà d'una correcta elecció del valor de les variables es pot millorar la seguretat i durabilitat del pont amb un xicotet increment del cost. A més, esta metodologia és aplicable a qualsevol tipus d'estructura i material. / García Segura, T. (2016). Efficient design of post-tensioned concrete box-girder road bridges based on sustainable multi-objective criteria [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/73147

Since the discovery of superconductivity, its fantastic properties have fascinated the scientific community. The discovery of high critical temperature (Tc) superconducting compositions further inspires the wide applications of superconductors with relatively inexpensive liquid nitrogen cooling. Recently, the integration of superconductivity and optical waveguides has put forward the potential for ultrasensitive, ultra-fast and ultralow noise light detectors. However, simple and cost effective superconductor designs and fabrication processes are still required to enable wide implementation. The objective of this research was to study the fabrication of the superconductor core fibers with a fused silica cladding via the melt-draw approach, as well as develop appropriate characterization techniques to describe the fibers produced. In addition, a further objective was to determine the cooling efficiency of ordered holes around a superconductor core and construction of a one dimensional (1-D) single-phase steady state model to predict the heat transfer during cryogenic liquid transfer inside glass tube. In this thesis, both Pb and YBCO superconductor core fibers with fused silica cladding have been demonstrated. The fibers were fabricated via the melt-draw technique and maintained overall diameters ranging from 200-900 μm and core diameters of 100-800 μm. Surface morphology, chemical composition, interface effect, and superconductivity were further investigated. Surface morphology analysis confirmed that the Pb and YBCO core fibers possessed good circularity and clean interfaces between the core and cladding. Both the Pb and YBCO cores were relatively dense after the melt-draw process. The melt-draw process avoided contamination during fabrication as indicated by the composition analysis. Limited PbO was examined on the Pb core surface but further action will be required to detect the source of oxygen. The YBCO core maintained a stoichiometric ratio comparable to the superconducting phase even after the melt forming process. The elemental mapping showed that limited cross-diffusion occurred between the Pb core and fused silica cladding. Conversely significantly more elemental cross interaction between the core and cladding was noted for the YBCO core fiber. Superconductivity of the Pb core was verified by a custom designed four-probe technique in liquid helium. The YBCO core was also confirmed to be superconductive after heat treatment with O₂ present. The feasibility of efficient cooling by the holey glass tubes was confirmed. A 1-D single-phase steady state model was constructed to evaluate the heat transfer mechanism. The experimental results are in reasonable agreement to the theoretical calculation. / Master of Science

State of the art filter bank structures achieve practically perfect reconstruction with very high computational efficiency. However, the increase in computational requirements due to the need to process increasingly wider band signals is paramount. New filter bank structures that provide extra information about a signal while achieving the same level of required efficiency, and perfect reconstruction properties, need to be developed. In this work a new filter bank structure, the interpolated FIR (IFIR) filter bank is developed. Such a structure combines the concepts of filter banks, and interpolated FIR filters. The filter design procedures for the IFIR filter bank are developed and explained. The resulting structure was compared with the non-maximally-decimated filter bank (NMDFB), achieving the same performance in terms of the number of multiplications required per sample and the overall distortion introduced by the system, when operating with Nyquist prototype filters. In addition, the IFIR filter is tested in both simulated and real communication environments. Performance, in terms of bit-error-rate, was found to not be degraded significantly when using the IFIR filter bank system for transmission and reception of QPSK symbols. / Master of Science

Earlier designs were driven by the goal of achieving higher performance, but lately, energy efficiency has emerged as an even more important design principle. Strong demand from the consumer electronics drives research in the low power and energy-efficient methodologies. Moreover, with exponential increase in the number of transistors on a chip and with further technology scaling, variability in the design is now of greater concern. Variations can make the design unreliable or the design may suffer from sub-optimal performance. Through the work in this thesis, we present a multi-dimensional investigation into the design of variation aware energy-efficient systems. Our overarching methodology is to use system-level decisions to mitigate undesired effects originating from device-level and circuit-level issues. We first look into the impact of process variation (PV) on energy efficient, scalable throughput many-core DSP systems. In our proposed methodology, we leverage the benefits of aggressive voltage scaling (VS) for obtaining energy efficiency while compensating for the loss in performance by exploiting parallelism present in various DSP designs. We demonstrate this proposed methodology consumes 8% - 77% less power as compared to simple dynamic VS over different workload environments. Later, we show judicious system-level decisions, namely, number of cores, and their operating voltage can greatly mitigate the effects of PV and consequently, improve the energy efficiency of the design. We also present our analysis discussing the impact of aging on the proposed methodology. To validate our proposed system-level approach, design details of a prototype chip fabricated in the 90nm technology node and its findings are also presented. The chip consists of 8 homogeneous FIR cores, which are capable of running from near-threshold to nominal voltages. In the 20-chip population, we observe 7% variation in the speed at nominal voltage (0.9V) and 26% at near threshold voltage (0.55V) among all the cores. We also observe 54% variation in power consumption characteristics of the cores. The chip measurement results show that our proposed methodology of judiciously selecting the cores and their operating voltage can result in 6.27% - 28.15% more energy savings for various workload environments, as compared to globally voltage scaled systems. Furthermore, we present the impact of temperature variations on the energy-efficiency of the above systems. We also study the problem of voltage variations in the integrated circuits. We first present the characteristics of a dynamic voltage noise as measured on a 28nm FPGA. We propose a fully digital on-chip sensor that can detect the fast voltage transients and alert the system of voltage emergency. A traditional approach to mitigate this problem is to use safety guardbands. We demonstrate that our proposed sensor system will be 6% - 27.5% more power efficient than the traditional approach. / Ph. D.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第24146号 / 工博第5033号 / 新制||工||1786(附属図書館) / 京都大学大学院工学研究科航空宇宙工学専攻 / (主査)教授 藤本 健治, 教授 松野 文俊, 教授 森本 淳 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM