With the fast growth of data-intensive distributed systems today, more novel and principled approaches are needed to improve the system efficiency, ensure the service quality to satisfy the user requirements, and lower the system running cost. This dissertation studies the design issues in the data-intensive distributed systems, which are differentiated from other systems by the heavy workload of data movement and are characterized by the fact that the destination of each data flow is limited to a subset of available locations, such as those servers holding the requested data. Besides, even among the feasible subset, different locations may result in different performance. The studies in this dissertation improve the data-intensive systems by exploiting the data storage location flexibility. It addresses how to reasonably determine the data placement based on the measured request patterns, to improve a series of performance metrics, such as the data access latency, system throughput and various costs, by the proposed hypergraph models for data placement. To implement the proposal with a lower overhead, a sketch-based data placement scheme is presented, which constructs the sparsified hypergraph under a distributed and streaming-based system model, achieving a good approximation on the performance improvement. As the network can potentially become the bottleneck of distributed data-intensive systems due to the frequent data movement among storage nodes, the online data placement by reinforcement learning is proposed which intelligently determines the storage locations of each data item at the moment that the item is going to be written or updated, with the joint-awareness of network conditions and request patterns. Meanwhile, noticing that distributed memory caches are effective measures in lowering the workload to the backend storage systems, the auto-scaling of memory cache clusters is studied, which tries to balance the energy cost of the service and the performance ensured. As the outcome of this dissertation, the designed schemes and methods essentially help to improve the running efficiency of data-intensive distributed systems. Therefore, they can either help to improve the user-perceived service quality under the same level of system resource investment, or help to lower the monetary expense and energy consumption in maintaining the system under the same performance standard. From the two perspectives, both the end users and the system providers could obtain benefits from the results of the studies. / Graduate

Tripod, an integrated bi-level transportation management system, is a smartphone application from the potential user’s point of view which would be instantly accessed prior to performing the trip. Since there are constantly several alternatives for any trip, Tripod considers a series and combination of various parameters, including departure time, mode and route, and rewards for each alternative with a number of redeemable points for goods and services called “Tokens”. The framework responsible for computing the optimized number of tokens awarded to the set of available alternatives in order to minimize the system-wide energy consumption under a constrained Token budget, is the System Optimization (SO) implemented in Tripod. To do so, a higher number of tokens would be awarded to the alternatives, guaranteeing a larger energy saving, less energy consumption, alternatively. SO is multimodal whereby public transit, private car, carpooling, etc. are being considered as the potential travel modes. Furthermore, SO is dynamic, predictive and personalized: the same alternative is rewarded differently, depending on the current and predicted future condition of the network and on the individual’s profile. In order to solve this problem, a multimodal simulation-based optimization model will be elaborated.

This thesis presents research focused on developing statistical methods with emphasis on techniques that can be used for the analysis of data in imaging studies and sparse estimations for applications in high-dimensional data. The first contribution addresses the pixel/voxel-labeling problem for spatial hidden Markov models in image analysis. We formulate a Gaussian spatial mixture model with Potts model used as a prior for mixture allocations for the latent states in the model. Jointly estimating the model parameters, the discrete state variables and the number of states (number of mixture components) is recognized as a difficult combinatorial optimization. To overcome drawbacks associated with local algorithms, we implement and make comparisons between iterated conditional modes (ICM), simulated annealing (SA) and hybrid ICM with ant colony system (ACS-ICM) optimization for pixel labelling, parameter estimation and mixture component estimation. In the second contribution, we develop ACS-ICM algorithm for spatiotemporal modeling of combined MEG/EEG data for computing estimates of the neural source activity. We consider a Bayesian finite spatial mixture model with a Potts model as a spatial prior and implement the ACS-ICM for simultaneous point estimation and model selection for the number of mixture components. Our approach is evaluated using simulation studies and an application examining the visual response to scrambled faces. In addition, we develop a nonparametric bootstrap for interval estimation to account for uncertainty in the point estimates. In the third contribution, we present sparse estimation strategies in linear mixed model (LMM) for longitudinal data. We address the problem of estimating the fixed effects parameters of the LMM when the model is sparse and predictors are correlated. We propose and derive the asymptotic properties of the pretest and shrinkage estimation strategies. Simulation studies is performed to compare the numerical performance of the Lasso and adaptive Lasso estimators with the pretest and shrinkage ridge estimators. The methodology is evaluated through an application of a high-dimensional data examining effective brain connectivity and genetics. In the fourth and final contribution, we conduct an imaging genetics study to explore how effective brain connectivity in the default mode network (DMN) may be related to genetics within the context of Alzheimer’s disease. We develop an analysis of longitudinal resting-state functional magnetic resonance imaging (rs-fMRI) and genetic data obtained from a sample of 111 subjects with a total of 319 rs-fMRI scans from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. A Dynamic Causal Model (DCM) is fit to the rs-fMRI scans to estimate effective brain connectivity within the DMN and related to a set of single nucleotide polymorphisms (SNPs) contained in an empirical disease-constrained set. We relate longitudinal effective brain connectivity estimated using spectral DCM to SNPs using both linear mixed effect (LME) models as well as function-on-scalar regression (FSR). / Graduate

According to non-stopable progress in wireless communications, it is desirable to integrate the RF front-end with the baseband building blocks of communication circuits into a one chip in the recent years. The CMOS technology advances, this is the reason why it becomes attractive for system-on-a-chip implementation, but CMOS device is getting shrink, so the channel electric field increasing and the hot carrier (HCI) effect becomes more significant. If the oxide is scaled down to less than 3 nm, then there is the possibility of soft or hard oxide breakdown (S/HBD) often takes place. As a result of the oxide trapping and interface generation is the long term performance drift and related reliability problems in devices and circuits. During the scaling and increasing chip power dissipation operating temperatures for device have also is increasing. Another reliability concern is the negative bias temperature instability (NBTI) caused by the interface traps under high temperature and negative gate voltage bias are arising while the operation temperature of devices is increasing. Parameter’s extraction is a very important part of the current electronic components modeling process, as it looking for the value of the unknown parameters in mathematical model, which represents physical behavior of given electronic component. The problem of parameter extraction is that fits electronic components mathematical model to a measured data set is an ill-posed problem and its solution is inherently difficult. This diploma thesis presents the parameter extraction, optimization methodology and verifies it on a case study of a MOSFET mathematical models (LEVEL1, LEVEL2 and LEVEL3) parameter extraction. The presented nonlinear method is based on the method of the least squares, which is solved with the aid of Levenberg- Marquardt’s algorithm.

In Kombination mit erneuerbaren Energien (EEG) werden Elektrofahrzeuge (EVs) als wichtiger Bestandteil einer Transformation hin zu nachhaltigen Energiesystemen angesehen. Obwohl EVs heute nur einen geringen Anteil an der Fahrzeugdurchdringung in Deutschland darstellen, ist es das Ziel der Bundesregierung, dass im Jahr 2030 sechs Millionen EVs auf deutschen Straßen fahren sollen. Die Realisierung dessen hätte aufgrund des daraus resultierenden zusätzlichen Strombedarfs erhebliche Auswirkungen auf das Stromsystem. Wie hoch diese sind, hängt maßgeblich von der Ladestrategie der Fahrzeuge ab und ist der Forschungsgegenstand dieser Arbeit. Die übergeordnete ökonomische Fragestellung lautet: Welche Auswirkungen haben unterschiedliche EV-Ladestrategien auf Strommengen und -preise in einem Stromsystem mit einem hohen Anteil an erneuerbaren Energien? Zur Beantwortung dessen wird zunächst der zeitabhängige Strombedarf von EVs bewertet. Im Anschluss, werden die EV-Ladestrategien unkontrolliertes Laden (UNC), kostengesteuertes Laden (DSM) und bidirektionales Laden (V2G) in einem europäischen Strommarktmodell umgesetzt und die Auswirkungen quantifiziert. Dadurch wurden folgende Erkenntnisse erlangt: EVs tragen zu einer besseren Integration der EEG bei, da alle drei Ladestrategien deren Abregelung reduzieren. Der zusätzliche Spitzenlastbedarf aufgrund von UNC wird je Millionen EVs im schlimmsten Fall auf 560 MW geschätzt. Entsprechend des Fahrverhaltens variiert die Stromnachfrage stark zwischen Werktagen und Wochenendtagen. An Werktagen sind die Spitzenwerte fast dreimal so hoch wie an Wochenendtagen. Wird durch UNC die Stromnachfrage erhöht, bedarf es des vermehrten Einsatzes von Spitzenlastkraftwerken, was zu steigenden Preisspitzen führt. Im Gegensatz dazu verschieben die beiden flexiblen Ladestrategien DSM und V2G die EV-Stromnachfrage in Zeiten mit geringer residualer Netzlast bzw. bei V2G deutlich zugunsten von Kraftwerken mit den niedrigsten Grenzkosten. Dies führt bei DSM zu einer Anhebung der Preise in Schwachlastzeiten. Bei V2G wird die Preisstruktur erheblich geglättet, indem Spitzenlastpreise reduziert und Schwachlastpreise deutlich erhöht werden. An Wochenenden ist dieser Effekt bei V2G noch stärker als an Werktagen, da ein großer Teil der EVs als stationärer Speicher genutzt werden kann. Neben ökonomischer Effizienz hat dies teilweise unerwünschte ökologische Nebenwirkungen. So werden im Fall von V2G bei niedrigen CO2-Preisen emissionsintensive Technologien wie Braunkohlekraftwerke begünstigt. Nichtsdestotrotz führen systemische Effekte, nämlich die Reduzierung von EEG-Abschaltungen, die Substitution von Spitzenlastkraftwerken und ein erhöhter Stromaustausch mit den Nachbarländern zu einer Gesamtreduktion der CO2-Emissionen. Bei hohen CO2-Preisen sind die Effekte durch V2G hinsichtlich der CO2-Emissionen und der ökonomischen Effizienz durchweg positiv. Begrenzt werden diese Vorteile von V2G durch wirtschaftliche Sättigungseffekte, welche bereits ab zwei Millionen Fahrzeugen deutlich werden. / In combination with renewable energy sources (RES), electric vehicles (EVs) are seen as an important element of a transformation towards sustainable energy systems. Although EVs currently represent only a small fraction of vehicle penetration in Germany, it is the goal of the German government to have six million EVs on German roads by 2030. The achievement of this would have a significant impact on the electricity system due to the resulting additional energy demand. How large these impacts are is the subject of this work. The overarching economic research question is: What effects do different EV charging strategies have on quantities and prices in a power system with a high share of RES? To answer this question, the time-dependent electricity demand of EVs is initially evaluated. Subsequently, the EV charging strategies uncontrolled charging (UNC), demand side management (DSM), in the sense of cost effective charging and bidirectional charging, i.e. vehicle-to-grid (V2G) are implemented in a European electricity market model and the impacts quantified. To summarize the findings: EVs contribute to the integration of RES, since all three charging strategies reduce curtailment. In the worst case scenario, the additional peak load demand due to UNC is estimated at 560 MW per million EVs. The demand for electricity varies greatly between working days and weekend days, depending on the driving patterns. On working days, the peak demand is almost three times as high as on weekend days. Overall, UNC leads to the increased use of peak load power plants, which leads to rising price peaks. In contrast, the two flexible charging strategies DSM and V2G shift the EVs' electricity demand in times of low residual grid load or, in the case of V2G, significantly in favour of the power plants with the lowest marginal costs. With DSM, this results in an increase in prices during off-peak periods. With V2G, the price structure is considerably smoothed by reducing peak load prices and significantly increasing off-peak prices. On weekend days this effect is even stronger with V2G than on working days, since a large part of the EVs can be used as stationary storage. In addition to economic effciency, this has partly undesirable ecological side effects. In the case of V2G, emission-intensive technologies such as lignite-fired power plants are promoted at low CO2 prices. Nevertheless, systemic effects, namely the reduction of RES curtailment, the substitution of peak load power plants, and an increased electricity exchange with neighboring countries, lead to an overall reduction of the CO2 emissions. These benefits of V2G are limited due to economic saturation effects, which are already noticeable starting at two million vehicles.

As Sweden switches to increasing renewable electricity production the demand on the energy grid and energy market will become higher. Since a bigger part of the electricity consumption will come from flowing energy sources the production will become less stable and harder to plan with the consumption. The inertia of the electrical system will also decrease since solar and wind power are not synchronously connected to the electrical system which will make the system more sensitive to interference. In order to keep the short-term balance so that the frequency remains at 50 𝐻𝑧, Svenska kraftnät has several reserves at their disposal. As of summer 2020, Svenska kraftnät will launch a new reserve called Fast frequency reserve (FFR) with the purpose to deal with rapid imbalances. By supplementing a wind farm with a battery energy store system (BESS), it becomes possible to even out the wind farm's intermittent electricity production by applying peak shaving and lower the grid costs for the wind farm. Because a BESS can provide power within a fraction of a second and is therefore is suitable to provide FFR. To study the profitability and determine what capacity and power a BESS needs for peak shaving and FFR with a wind farm, an optimization model was built in MATLAB to study the profitability of a BESS with multiple power and capacity combination. In addition, the cycling of the BESS and the limitation of peak shaving was also studied to get deeper knowledge about the limitations. The optimization model is using hourly generation data from a wind farm in northern Sweden. Besides the BESS optimization, a separate optimization model was built in order regulate the output power to minimize the generation cost by prolonging the service life of a wind turbine (WTG). The purpose of this optimization was to study if regulating the output power could lower the generation cost, more for the WTG. In addition of the net income the loss of electricity was also studied. The optimization used hourly data during one time period every season during of 2019. The optimization for the BESS showed that the levelized cost of storage (𝐿𝐶𝑂𝑆𝐸) is currently too high for a BESS to be used for only peak shaving with a wind farm. For a BESS to be feasible together with a wind farm the 𝐿𝐶𝑂𝑆𝐸 needs to decrease towards 𝐿𝐶𝑂𝑆𝐸<6 𝐸𝑈𝑅/𝑀𝑊ℎ, and when the BESS also supplied FFR the income increased between 1.5 – 8% depending on the power output for the BESS. The capacity was the limiting factor for the BESS when preforming peak shaving while FFR was limited by the power because of the low energy demand in FFR. Lowering the power output for a WTG resulted in an increased net income for every month between 10 – 90% although this increased income will become more apparent when the operation and maintenance cost starts to drop over a couple of year but this open up a discussion of how an owner should operate there WTG.

The following thesis looks into modeling a chilled water system equipped with variable speed drives on different piece of equipment and optimization of system setpoints to achieve energy savings. The research was done by collecting data from a case-study and developing a system of component models that could be linked to simulate the overall system operation.

Electrical energy is the basic necessity for the economic development of human societies. In recent decades, the electricity industry is undergoing enormous changes, which have evolved into a large-scale and competitive industry. The integration of volatile renewable energy, and the emergence of transmission switching (TS) techniques bring great challenges to the existing power system operations problems, especially security-constrained unit commitment (SCUC) solution engines. In order to deal with the uncertainty of volatile renewable energy, scenario-based stochastic optimization approach has been widely employed to ensure the reliability and economic of power systems, in which each scenario would represent a possible system situation. Meanwhile, the emergence of TS techniques allows the system operators to change the topology of transmission systems in order to improve economic benefits by mitigating transmission congestion. However, with the introduction of extra scenarios and decision variables, the complexity of the SCUC model increases dramatically and more computational efforts are required, which might make the power system operation problems difficult to solve and even intractable. Therefore, an advanced solution technique is urgently needed to solve both stochastic SCUC problems and TS-based SCUC problems in an effective and fast way. In this dissertation, a decomposition framework is presented for the optimal operation of the large-scale power system, which decomposes the original large-size power system optimization problem into smaller-size and tractable subproblems, and solves these decomposed subproblems in a parallel manner with the help of high performance computing techniques. Numerical case studies on a modified I 118-bus system and a practical 1168-bus system demonstrate the effectiveness and efficiency of the proposed approach which will offer the power system a secure and economic operation under various uncertainties and contingencies.

Municipalities globally are recognizing their role in mitigating climate change and are actively working to reduce carbon emissions. This complex challenge is heightened in areas like Northern Sweden, where municipalities are adapting to accommodate new industries essential for meeting global climate targets, subsequently changing the energy landscape. The local administration must not only decarbonize existing energy use but also develop infrastructure for the new industries, all while fostering sustainable and appealing cities where residents aspire to live. However, the trajectory of these changes and the subsequent future energy requirements remain uncertain. This study aims to assist the local administration in navigating through these uncertainties and setting ambitious climate and energy targets aligned with the goals of the Paris Agreement and sustainable developments. The research explores how model based scenario analysis can be improved to identify a set of relevant pathways that the municipalities can adopt by employing system analysis, energy system optimization, and scenario analysis. The study focuses on Gällivare municipality in Northern Sweden and employs the TIMES-City model to develop the energy system model of the municipality (RQ1). To identify relevant scenarios for local energy transition a framework for developing "Glocal" scenarios has been established (RQ2). These glocal scenarios incorporate global, national, and local socioeconomic trends into a coherent narrative and provide a more holistic and realistic view of potential future pathways (Paper 2). Additionally, a set of SDG indicators for evaluating the sustainability of different scenarios has been developed and applied in the model (RQ3, Paper 3). While the study focuses on Gällivare, the "glocal" scenario framework and SDG indicators developed in this research can be utilized by municipalities across the globe for identifying their climate and energy targets.

The diploma theses Optimization of logistics system in the Food Bank of South Bohemia, is focused on design of optimimal logistics system in the Food Bank of South Bohemia. Social benefits of food banks in general consist in solidarity and reducing of food waste. As nonprofit organizations the food bank is supported by volunteers, often students who are looking for savings in contemporary logistic processes. This thesis is focused on finding optimization in the whole contemporary system of logistics based on solving current issues in the field of distribution, transport and storage system of the Food Bank.