京都大学 / 新制・課程博士 / 博士(工学) / 甲第23884号 / 工博第4971号 / 新制||工||1776(附属図書館) / 京都大学大学院工学研究科機械理工学専攻 / (主査)教授 小森 雅晴, 教授 松野 文俊, 教授 藤本 健治 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

High frame rate video (HFV) is an important investigational tool in sciences, engineering and military. In ultrahigh speed imaging, the obtainable temporal, spatial and spectral resolutions are limited by the sustainable throughput of in-camera mass memory, the lower bound of exposure time, and illumination conditions. In order to break these bottlenecks, we propose a new coded video acquisition framework that employs K>1 cameras, each of which makes random measurements of the video signal in both temporal and spatial domains. For each of the K cameras, this multi-camera strategy greatly relaxes the stringent requirements in memory speed, shutter speed, and illumination strength. The recovery of HFV from these random measurements is posed and solved as a large scale l1 minimization problem by exploiting joint temporal and spatial sparsities of the 3D signal. Three coded video acquisition techniques of varied trade o s between performance and hardware complexity are developed: frame-wise coded acquisition, pixel-wise coded acquisition, and column-row-wise coded acquisition. The performances of these techniques are analyzed in relation to the sparsity of the underlying video signal. To make ultra high speed cameras of coded exposure more practical and a fordable, we develop a coded exposure video/image acquisition system by an innovative assembling of multiple rolling shutter cameras. Each of the constituent rolling shutter cameras adopts a random pixel read-out mechanism by simply changing the read out order of pixel rows from sequential to random. Simulations of these new image/video coded acquisition techniques are carried out and experimental results are reported. / Dissertation / Doctor of Philosophy (PhD)

A growing need for clean and sustainable energy is causing a significant shift in the electricity generation paradigm. In the electricity system of the future, integration of renewable energy sources with smart grid technologies can lead to potentially huge economical and environmental benefits ranging from lesser dependency on fossil fuels and improved efficiency to greater reliability and eventually reduced cost of electricity. In this context, microgrids serve as one of the main components of smart grids with high penetration of renewable resources and modern control strategies. This dissertation is concerned with developing optimal control strategies to manage an energy storage unit in a grid-connected microgrid under uncertainty of electricity demand and prices. Two methods are proposed based on the concept of rolling horizon control, where charge/discharge activities of the storage unit are determined by repeatedly solving an optimization problem over a moving control window. The predicted values of the microgrid net electricity demand and electricity prices over the control horizon are assumed uncertain. The first formulation of the control is based on the scenario-based stochastic conditional value at risk (CVaR) optimization, where the cost function includes electricity usage cost, battery operation costs, and grid signal smoothing objectives. Gaussian uncertainty is assumed in both net demand and electricity prices. The second formulation reduces the computations by taking a worst-case CVaR stochastic optimization approach. In this case, the uncertainty in demand is still stochastic but the problem constraints are made robust with respect to price changes in a given range. The optimization problems are initially formulated as mixed integer linear programs (MILP), which are non-convex. Later, reformulations of the optimization problems into convex linear programs are presented, which are easier and faster to solve. Simulation results under different operation scenarios are presented to demonstrate the effectiveness of the proposed methods. Finally, the energy management problem in network of grid-connected microgrids is investigated and a strategy is devised to allocate the resulting net savings/costs of operation of the microgrids to the individual microgrids. In the proposed approach, the energy management problem is formulated in a deterministic co-operative game theoretic framework for a group of connected microgrids as a single entity and the individual savings are distributed based on the Shapley value theory. Simulation results demonstrate that this co-operation leads to higher economical return for individual microgrids compared to the case where each of them is operating independently. Furthermore, this reduces the dependency of the microgrids on the utility grid by exchanging power locally. / Thesis / Master of Applied Science (MASc)

Energy management strategies for microgrids, containing energy storage, renewable energy sources (RES), and electric vehicles (EVs); which interact with the grid on an individual basis; are presented in Chapter 3. An optimization problem to reduce cost, formulated over a rolling time horizon, using predicted values of load demand, EV connection/disconnection times, and charge levels at time of connection, is described. The solution provides the on-site storage and EV charge/discharge powers. For the first time, both bidirectional and unidirectional charging are considered for EVs and a controller which accommodates uncertainties in EV energy levels and connection/disconnection times is presented. In Chapter 4, a stochastic chance constraints based optimization is described. It affords significant improvement in robustness, over the conventional controller, to uncertainties in system parameters. Simulation results demonstrate that the stochastic controller is at least twice as effective at meeting the desired EV charge level at specific times compared to the non-stochastic version, in the presence of uncertainties. In Chapter 5, a network of microgrids, containing RES and batteries, which trade energy among themselves and with the utility grid is considered. A novel distributed energy management system (EMS), based on a central EMS using a Multi-Objective (MO) Rolling Horizon (RH) scheme, is presented. It uses Alternating Direction Method of Multipliers (ADMM) and Quadratic Programming (QP). It is inherently more data-secure and resilient to communication issues than the central EMS. It is shown that using an EMS in the network provides significant economic benefits over MGs connected directly to the grid. Simulations demonstrate that the distributed scheme produced solutions which are very close to those of the central EMS. Simulation results also reveal that the faster, less memory intensive distributed scheme is scalable to larger networks -- more than 1000 microgrids as opposed to a few hundreds for the central EMS. / Thesis / Doctor of Philosophy (PhD)

Die Reduzierung von Ausschuss und unnötiger Nacharbeit ist ein elementares Ziel der Fertigungsindustrie. Mit der zunehmenden Datenverfügbarkeit und den Entwicklungen auf dem Gebiet der künstlichen Intelligenz (KI) für industrielle Anwendungen, wird auch im Bereich des Radial-Axial Ringwalzens (RAW) der Einsatz des maschinellen Lernens (ML) eruiert. Die Anwendungen hier sind beispielsweise die Prozessauslegung oder die Vorhersage der Ringqualität [1]. Allerdings ist die Genauigkeit dieser Vorhersagen derzeit noch durch die Menge und Qualität der Daten begrenzt [2]. Um das überwachte Lernen zur Vorhersage der Ringqualität anzuwenden, muss eine umfangreiche Datenbasis für Gut- und Ausschussteile erzeugt werden. Eine Möglichkeit, bestehende Datenbasen zu erweitern, besteht in der Nutzung von Prozesssimulationen zur Generierung synthetischer Daten. Im Bereich des Warmringwalzens gibt es jedoch derzeit keine schnelle Simulationsmethode, mit der eine ausreichend große synthetische Datenbank von gewalzten Teilen mit Form- oder Prozessfehlern generiert werden kann. Die Forschung zum Transferlernen zwischen verschiedenen Walzwerken und Datensätzen hat die neuartige Idee hervorgebracht, das Kaltringwalzen als Untersuchungsgegenstand heranzuziehen [2]. Im Folgenden wird untersucht, inwieweit das Kaltringwalzen, als ähnlicher Prozess, für die zukünftige Übertragung von Modellen und Ergebnissen auf das RAW verwendet werden kann. Im Vergleich zum RAW wird die Umformung beim Kaltringwalzen nur durch zwei Radialwalzen erreicht und der Prozess wird bei Raumtemperatur durchgeführt. Diese vereinfachte Verfahrensweise erlaubt es, ein halb-analytisches Modell zu entwickeln, das im Vergleich zu herkömmlichen FEM-Ansätzen, bei akzeptabler Genauigkeit, viel weniger Berechnungszeit erfordert. Zudem ermöglichen die geringere Ringgröße und der einfachere Walzprozess die Durchführung umfangreicher Forschungswalzungen zur Überprüfung der Qualität der synthetischen Daten.

Reducing scrap products and unnecessary rework has always been a goal of the manufacturing industry. With the increasing data availability and the developments in the field of artificial intelligence (AI) for industrial applications, machine learning (ML) has been applied to radial-axial ring rolling (RARR) to predict product quality [1]. However, the accuracy of these predictions is currently still limited by the quantity and quality of the data [2]. In order to apply supervised learning to predict part quality and possible scrap parts, there must be plenty of datasets logged for both good and scrap parts. One suitable way to increase the number of datasets is to utilize simulation strategies to generate synthetic datasets. However, in the hot ring rolling field, there is no fast simulation method that can be used to generate a sufficiently large synthetic database of rolled parts with form or process errors. The research on transfer learning between different mills and datasets has offered a new idea of taking a cold ring rolling process as the object of study [2]. Next it will investigate the extent to which the cold ring rolling can be used as a similar process for future transfer of models and results to radial-axial ring rolling. Compared to RARR, the cold ring rolling is a process under room temperature and contains complete radial forming instead of simultaneous forming in the radial and axial directions. The simpler forming mechanism makes it possible to build a semi-analytical model, which takes much less time compared to conventional FEMapproaches under acceptable accuracies. Furthermore, the smaller ring geometry, simplified rolling process and reduced energy consumption mean that in-house experiments can be conducted to verify the quality of the synthetic data based on confidence intervals.

Anwendung von maschinellen Lernverfahren (ML) in der Produktionstechnik, in Zeiten der Industrie 4.0, stark angestiegen. Insbesondere die Datenverfügbarkeit ist an dieser Stelle elementar und für die erfolgreiche Umsetzung einer ML-Applikation Voraussetzung. Falls für eine gegebene Problemstellung die Datenmenge oder -qualität nicht ausreichend ist, können Techniken, wie die Datenaugmentierung, der Einsatz von synthetischen Daten sowie das Transferlernen von ähnlichen Datensätzen Abhilfe schaffen. Innerhalb dieser Ausarbeitung wird das Konzept des Transferlernens im Bereich das Radial-Axial Ringwalzens (RAW) angewendet und am Beispiel der Zeitreihenprognose des Außendurchmessers über die Prozesszeit durchgeführt. Das Radial-Axial Ringwalzen ist ein warmumformendes Verfahren und dient der nahtlosen Ringherstellung.

Due to the increasing computing power and corresponding algorithms, the use of machine learning (ML) in production technology has risen sharply in the age of Industry 4.0. Data availability in particular is fundamental at this point and a prerequisite for the successful implementation of a ML application. If the quantity or quality of data is insufficient for a given problem, techniques such as data augmentation, the use of synthetic data and transfer learning of similar data sets can provide a remedy. In this paper, the concept of transfer learning is applied in the field of radial-axial ring rolling (rarr) and implemented using the example of time series prediction of the outer diameter over the process time. Radial-axial ring rolling is a hot forming process and is used for seamless ring production.

Modern automotive systems increasingly depend on camera sensors to gather safetycriticaldata used in driver-assisting features of the system. These features can consist offor example, lane-keeping assist and automatic braking where the sensors register objectswithin certain distances. When these camera sensors gather information, the time of theimage is critical for the calculation of speeds, distances, and size of any potential registeredobject in the frame. Limitations of bandwidth and computing in such vehicles creates aneed to use special cameras that do not capture the whole image simultaneously but insteadcapture the images piecewise. These cameras are called rolling shutter cameras. Thisputs pressure on defining when an image was captured when different parts of the imagewere captured at different points in time. For this thesis, this point in time is defined as thechronological middle point in between the camera starting to capture an image and when ithas collected the final part of it. This thesis performs a mapping-study to evaluate methodsto verify the timestamp of an image generated from rolling shutter cameras. Further, thisthesis proposes a new method using multiple digital clocks and presents its performanceusing a proof-of-concept implementation to prove the method’s ability to accurately representtime with sub-millisecond accuracy.

No description available.