Search

End-to-end deep learning approaches has been proven to be efficient in autonomous driving and robotics. By using deep learning techniques for decision-making, those systems are often referred to as a black box, and the result is driven by data. In this paper, we propose PaaS (Planning as a Service), a vanilla module to generate local trajectory planning for autonomous driving in CARLA simulation. Our method is submitted in International CARLA Autonomous Driving Leaderboard (CADL), which is a platform to evaluate the driving proficiency of autonomous agents in realistic traffic scenarios. Our approach focuses on reactive planning in Frenet frame under complex urban street's constraints and driver's comfort. The planner generates a collection of feasible trajectories, leveraging heuristic cost functions with controllable driving style factor to choose the optimal-control path that satisfies safe travelling criteria. PaaS can provide sufficient solutions to handle well under challenging traffic situations in CADL. As the strict evaluation in CADL Map Track, our approach ranked 3rd out of 9 submissions regarding the measure of driving score. However, with the focus on minimizing the risk of maneuver and ensuring passenger safety, our figures corresponding to infraction penalty dominate the two leading submissions for 20 percent., final draft, submitting to ICSSE 2023

An approach to solving the problems of fracture mechanics at the scale level of continuum mechanics is proposed. For an elementary volume element, within the framework of this approach, the onset of the moment of formation of new material surfaces is considered. A statement and a method for solving a particular problem of fracture mechanics are proposed.

The problem of tension and shear of a thin isotropic plate weakened by a doubly periodic system of rectilinear cracks with end zones is considered. General representations of solutions are constructed that describes a class of problems with a doubly periodic stress distribution outside cracks. The main singular integral equation of the problem is reduced to an infinite system of algebraic equations. Due to the unknown size of the end zone, the system of equations turned out to be non-linear even with a linear bond deformation law.

Kovir capsule, a polyherbal medicine developed from Ren Shen Bai Du San formulation, has been used in various diseases including respiratory infections. A randomized, placebo-controlled, double-blind study was conducted to evaluate the safety and efficacy of Kovir capsule (TD0069) in the treatment of mild COVID-19 patients. Patients aged from 18 to 65 years who were PCR-confirmed with SARS-CoV-2 and had the mild disease were recruited and randomized to either Kovir capsule (34 patients) or placebo (32 patients) for up to 14 days or until discharge. Efficacy outcomes were time to viral clearance, daily viral load, time to symptom resolution, daily symptom score based on 16 pre-defined symptoms, and progression to severe/critical COVID-19. Safety outcomes were adverse events. Viral load decreased over time similarly in the two groups. Viral clearance time was also similar in both groups (median: 8 days). Kovir group had a more rapid decrease of symptom score and significantly lower time to symptom resolution than placebo (median: 4 vs. 7 days). Two patients in the placebo group developed severe COVID-19. No patient experienced adverse events. Kovir capsule is safe and can improve symptom resolution in mild COVID-19 patients. A large-scale trial is required to confirm these findings.

Worldwide, energy consumption and saving represent the main challenges for all sectors, most importantly in industrial and domestic sectors. The internet of things (IoT) is a new technology that establishes the core of Industry 4.0. The IoT enables the sharing of signals between devices and machines via the internet. Besides, the IoT system enables the utilization of artificial intelligence (AI) techniques to manage and control the signals between different machines based on intelligence decisions. The paper’s innovation is to introduce a deep learning and IoT based approach to control the operation of air conditioners in order to reduce energy consumption. To achieve such an ambitious target, we have proposed a deep learning-based people detection system utilizing the YOLOv3 algorithm to count the number of persons in a specific area. Accordingly, the operation of the air conditioners could be optimally managed in a smart building. Furthermore, the number of persons and the status of the air conditioners arepublished via the internet to the dashboard of the IoT platform. The proposed system enhances decision making about energy consumption. To affirm the efficacy and effectiveness of the proposed approach, intensive test scenarios are simulated in a specific smart building considering the existence of air conditioners. The simulation results emphasize that the proposed deep learning-based recognition algorithm can accurately detect the number of persons in the specified area, thanks to its ability to model highly non-linear relationships in data. The detection status can also be successfully published on the dashboard of the IoT platform. Another vital application of the proposed promising approach is in the remote management of diverse controllable devices.

We develop a physics based analytical model for AlGaN/GaN high electron mobility transistors (HEMT and MOS-HEMT) to study the I-V characteristics, current transfer characteristics, transconductance and drain-conductance. The model is modified from a model first presented by Chang and Fetterman for AlGaAs/GaAs HEMT (Chang and Fetterman in Solid-State Electron 30(5), 1987 [1], IEEE Trans Electron Dev Ed-34(I), 1987 [2]) The linear and non-linear drain currents have been separately calculated and merged them in order to evaluate the I-V characteristics. The threshold voltage has been evaluated from current transfer characteristics plot and verified from the transconductance parameter. We also consider the effect of polarization on threshold voltage of the devices. Moreover, to prove the exactness of our model, we compare our data with our experiment data for common HEMT, Hasan’s MOS-HEMT and Yoon’s HEMT with short channel.

The European 1 MW, 170 GHz industrial CW prototype gyrotron has been designed within EGYC (European GYrotron Consortium) in collaboration with the industrial partner Thales Electron Devices (TED) and under the coordination of Fusion for Energy (F4E). This is a conventional (hollow) cavity gyrotron that is based on the 1 MW, 170 GHz short-pulse (SP) modular gyrotron, which has been designed and manufactured by KIT in collaboration with TED. The SP prototype has been tested in multiple experimental campaigns since 2015 and the nominal cavity mode TE32,9 is exited at 170.1 GHz, producing RF power above 1 MW with 35 % interaction efficiency. The first phase of the experiments with the CW industrial gyrotron was successfully completed at KIT in 2016, verifying most of the ITER specifications. Short pulses (

This presentation reports on the 2019 status and focus areas in the development of an advanced gyrotron for the European DEMOnstration Fusion Power Plant (EU DEMO). As specified within the EUROfusion Work Package Heating and Current Drive (WP HCD), the requirements for the gyrotron are as follows: (i) ≥ 2 MW of RF output power at frequencies ≥ 200 GHz; (ii) a total efficiency ≥ 60 %; (iii) the options for multiple-purpose/multi-frequency operation and fast frequency step-tunability within a limited bandwidth of approximately ± 10 GHz. To meet those targets, the R&D activities are organized in five main branches: (1) the experimental verification of the advanced coaxial-cavity gyrotron technology at long pulses; (2) the development of an advanced coaxial-cavity gyrotron that meets the EU DEMO requirements; (3) the fundamental research on multi-stage depressed collectors for enhanced energy recovery; (4) the development of large broadband diamond windows to allow for fast frequency tunability of the gyrotron: (5) the studies on further innovations and improvements of critical gyrotron components and tools in view of optimization of performance, reliability, and industrialization. Possible exchange: Within EUROfusion the EU is working on specific gyrotrons for future EU DEMO. There is a collaboration possible. Additionally, the different institutions within EU are open for any discussion on possible collaboration with regards to tools developments, development of future high-power high-frequency gyrotron teststands, measurements technologies and prospective R&D of further advances in gyrotron technology. If the collaboration is performed within EUROfusion, then the topics will be in line with the approved WPHCD work program.

During the ongoing pre-concept phase (2014 – 2020) for a possible future European DEMOnstration Fusion Power Plant (DEMO) the activities within EUROfusion WP HCD EC Gyrotron R&D and Advanced Developments are focusing on options for near-term solutions, and, at the same time, on long-term even more advanced options. The near-term target for DEMO is to realize pulsed operation. According to the current baseline it will probably use an EC system operating at 170 GHz and 204 GHz is being assessed, whereas the long-term target aims for steady-state operation and frequencies for current drive up to 240 GHz. Common targets for both are an RF output power per unit of significantly above 1 MW (target: 2 MW) and a total gyrotron efficiency of significantly higher than 60 %. Frequency step-tunability and multi-purpose/multi-frequency operation have to be considered. Those targets shall be achieved by considering the coaxial-cavity gyrotron technology and advanced technologies for key components (e.g. CVD diamond-disk Brewster angle window). Advanced simulation and test tools are complementing the research and developments. Gyrotron development is additionally supported by a significant investment into a new multi-megawatt long-pulse gyrotron test stand which is under final installation at KIT currently.

___

In mobile communication environments, handover is a very important process to maintain mobile host’s connections to the network. Although a lot of researchers have tried to improve the handover performance, little deployment to satisfy the users can be seen in the real network due to the modification costs. In the current age, the existence of network with different radio access technologies has created a heterogeneous environment. Moreover, the mobile hosts available to users are becoming more sophisticated, which opens the opportunity for end-point approaches to alleviate the deployment issues. Furthermore, real-time applications such as VoIP and video conference are delay and quality sensitive, but not much research has considered qualitative requirement. Hence, a suitable technique is needed to alleviate the deployment issues as well as to cope with the stringent requirements of real-time applications. This paper presents a new end-point centric handover approach based on the ant colony probabilistic equation that utilizes the received signal strength and the mean opinion score to trigger the handover. Some simulations were conducted using OMNeT++ to measure the effectiveness of the proposed approach comparing with an existing method, the Endpoint Centric Handover (ECHO). The results show that the proposed approach has better performance compared to the ECHO.

Improvements of Heating & Current Drive (H&CD) systems are being investigated for a demonstration fusion power plant DEMO to deliver net electricity for the grid around 2050 [1]. Compared to ITER, which has to show the generation of 500 MW thermal power, the target of DEMO is the successful production of 300 to 500 MW electrical power to the grid and to aim for a self-sufficient Tritium fuel cycle [2]. Three H&CD systems are under development for DEMO in Europe, the Electron Cyclotron (EC) System, the Neutral Beam Injection (NBI) System and the Ion Cyclotron (IC) System. Based on present studies [3] for plasma ramp-up, ramp-down and flat top phases, to be further validated in more detailed simulations, the assumed total launched power needed from the H&CD system in DEMO is in the range of 50-100 MW, to be provided for plasma heating and control. Among other topics, the paper describes the new designs and R&D status of H&CD systems considered for their deployment in DEMO in Europe and the impact of the H&CD configurations on their performances based on those areas described in the European fusion electricity roadmap [4] for the integrated design and system development. These configurations encompass the operation of NBI with reduced Caesium consumption, with modular ion sources and improved injector wall-plug efficiencies, EC system with increased gyrotron frequencies above ITER ones, as well as multi-purpose and frequency step-tunable radio frequency sources, and related strategies for the fabrication of large Brewster angle diamond windows, together with EC launcher designs compatible with a fusion power plant environment, avoiding front-steering by using different launcher concepts. The project also elaborates on new solutions to further increase the wall-plug efficiencies of H&CD systems based on more advanced concepts, with the target to reduce the recirculating power fraction in future fusion power plants. Different studies under investigation will be discussed such as, for NBI, the photo-neutralization and, for EC, new concepts for gyrotron multi-stage depressed collector. This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. With special thanks to whole WPHCD team for the durable, constructive and fruitful collaboration. [1] Federici, G., et al., ´Overview of EU DEMO design and R&D activities´, Fus. Eng. Design, vol. 89, pp. 882-889, 2014 Wenninger, R., et al., 'The physics and technology basis entering European system code studies for DEMO', Nucl. Fusion 57, 016011, 2017 Vincenzi, P., et al, ´EU DEMO transient phases: main constraints and heating mix studies for rampup and ramp-down', SOFT 2016, accepted for publication in Fus. Eng. Design Romanelli, F., et al., ´Fusion electricity: a roadmap to the realisation of fusion energy´, EFDA, Nov. 2012

Zur Beherrschung der wachsenden Komplexität und zur Verkürzung der Entwicklungszeiten werden Steuer- und Regelsysteme in Fahrzeugen heute zunehmend modellbasiert entwickelt. Dabei werden Algorithmen als Blockdiagramme erstellt, welche aus Blöcken und Linien bestehen. Modelle können simuliert und so validiert werden. Darüber hinaus lassen sich aus den Modellen Programme für Steuergeräte in Fahrzeugen mithilfe von Codegeneratoren automatisch generieren. Hierbei hat sich MATLAB (MATrix LABoratory)/Simulink als De-facto-Standard etabliert. Bei der Erstellung und Überarbeitung von Simulink-Modellen wird die Modellstruktur häufig verändert und erweitert. Während es bereits Refactoring-Ansätze für objektorientierte Programmiersprachen wie Java und C# und für funktionale Programmiersprachen wie Haskell und Scala gibt, fehlt ein vergleichbares, integrales und durchgehendes Verfahren für Simulink-Modelle. Der erste Beitrag dieser Arbeit ist ein automatisiertes Verfahren zum Refactoring von Simulink-Modellen. Ein Bestandteil dieses Verfahrens ist ein Katalog von Operationen, welche für das Refactoring von Simulink-Modellen genutzt werden können. Zur Spezifikation von den Operationen sind sowohl eine Technik zur Analyse als auch eine Technik zur Transformation von Simulink-Modellen in einer einheitlichen Notation erarbeitet. Um Modelle während der Erstellung effizienter nachvollziehen zu können, insbesondere hinsichtlich der Signalflüsse und Schnittstellen, dienen in dieser Arbeit erweiterte Analyseoperationen. Mithilfe dieser Analyseoperationen ist es dem Modellierer möglich, Einsatzmöglichkeiten für Refactoring besser zu erkennen. Diese erweiterten Analyseoperationen sind der zweite Beitrag dieser Arbeit. Der letzte Beitrag dieser Arbeit ist eine Technik zur Erkennung und Handhabung von Model Smells nach der Modellerstellung. Als Model Smell wird ein Hinweis auf ein strukturelles Defizit bezeichnet, welches die Modellstruktur negativ beeinflusst. Zur Beseitigung von solchen Model Smells dienen die Refactoring-Operationen aus dem ersten Beitrag. Software has become the centerpiece of most control systems in cars. For the development of software for control systems, model-based software engineering (MBSE) has been widely employed, since it promises an increase in efficiency and quality of software development. In particular, the algorithm of a control system is usually modeled as dataflow-oriented block diagrams, called models. The models can be simulated to validate their behavior. Furthermore, software can be generated directly from the models by automated code generators. The tool MATLAB Simulink by The Mathworks has become a de facto standard for MBSE in the automotive domain. The creation and maintenance of a model typically require lots of structural changes to the model. In the world of object-oriented and functional programming languages, refactoring has become an established technique for improving the structural quality of code. With refactoring, the structure of a code body is improved through small restructuring steps without its external behavior being altered. Refactoring Simulink models by restructuring their structure is also a highly frequent activity. Today, due to the lack of tool support, the process of time-consuming and error-prone refactoring of Simulink models must be done manually. As a first contribution, this thesis describes a refactoring approach to automated refactoring of Simulink models. In particular, a catalog of refactoring operations are introduced that can be used to refactoring different aspects of a Simulink model such as structural hierarchy, signal structuring, interface and signal routing. As a second contribution, the thesis introduces several advanced analysis operations for Simulink models. The operations can be used to make signal flow, interface and interaction between parts of a model more explicit. With the help of these operations, the modeler can comprehend a complex model more efficiently and thus has a better of chance of spotting refactoring opportunities. As a last contribution, the thesis introduces a technique for detection and handling of model smells for Simulink models. Model smell is referred to as a indicator for a structural deficit that has a negative impact on the quality of the model structure. Refactoring operations are recommended for eliminating the model smells.

Background (BG) traffic generated automatically by applications or the operating system on devices is delay-insensitive traffic as it does not immediately affect user experiences. However, this traffic may significantly contend network resources with foreground (FG) traffic which conveys data/content to users. This contention directly affects the QoS and QoE of network services, specifically at the peak times. This paper proposes an integration approach to which delay-insensitive traffic is quickly detected and effectively transferred in accordance with the network conditions hence the best resources are yielded to FG traffic. For the traffic transferring, we propose two methods namely the router centric (RC) and the SDN-based approaches. In the RC method, the edge routers analyze network conditions, compute the suggested sending rates for individual traffic flows and send these values back to clients where the adjustment mechanisms are implemented. In the SDN-based method the course of actions for each traffic flow is computed by a remote controller. Experimental results reveal the effectiveness and the feasibility of the proposed approaches.

MathScheme is a long-range research project being conducted at McMaster University with the aim to develop a mechanized mathematics system in which formal deduction and symbolic computation are integrated from the lowest level. The novel notion of a biform theory that is a combination of an axiomatic theory and an algorithmic theory is used to integrate formal deduction and symbolic computation into a uniform theory. A major focus of the project has currently been on building a library of formalized mathematics called the MathScheme Library. The MathScheme Library is based on the little theories method in which a portion of mathematical knowledge is represented as a network of biform theories interconnected via theory morphisms. In this thesis, we describe a systematic explanation of the underlying techniques which have been used for the construction of the MathScheme Library. Then we describe several algebraic constructions that can derive new useful machinery by leveraging the information extracted from a theory. For instance, we show a construction that can reify the term algebra of a (possibly multi-sorted) theory as an inductive data type. Master of Science (MSc)

Conventional frequent pattern mining algorithms require users to specify some minimum support threshold. If that specified-value is large, users may lose interesting information. In contrast, a small minimum support threshold results in a huge set of frequent patterns that users may not be able to screen for useful knowledge. To solve this problem and make algorithms more user-friendly, an idea of mining the k-most interesting frequent patterns has been proposed. This idea is based upon an algorithm for mining frequent patterns without a minimum support threshold, but with a k number of highest frequency patterns. In this paper, we propose an explorative mining algorithm, called ExMiner, to mine k-most interesting (i.e. top-k) frequent patterns from large scale datasets effectively and efficiently. The ExMiner is then combined with the idea of “build once mine anytime” to mine top-k frequent patterns sequentially. Experiments on both synthetic and real data show that our proposed methods are more efficient compared to the existing ones.

Experimental plasma discharges in linear PANTA device are studied by Mach probe measurements, providing floating potential, ion saturation current, and parallel flow velocity time evolution, at different radii of the device. Spectral analysis indicates that drift waves and D'Angelo modes exist simultaneously in the plasma. A discrimination study shows they are located at different positions in radius and frequency. (C) 2017 The Japan Society of Plasma Science and Nuclear Fusion Research

Within the range of plasma shapes and plasma currents investigated, the electron confinement time, Tau_E increases with density, elongation and negative triangularity (-0.4

During the ongoing pre-concept phase (2014 – 2020) for a possible future European DEMOnstration Fusion Power Plant (DEMO) the activities within EUROfusion WP HCD EC Gyrotron R&D and Advanced Developments are focusing on options for near-term solutions, and, at the same time, on long-term even more advanced options. The near-term target for DEMO is to realize pulsed operation. According to the current baseline it will probably use an EC system operating at 170 GHz and 204 GHz is being assessed, whereas the long-term target aims for steady-state operation and frequencies for current drive up to 240 GHz. Common targets for both are an RF output power per unit of significantly above 1 MW (target: 2 MW) and a total gyrotron efficiency of significantly higher than 60 %. Frequency step-tunability and multi-purpose/multi-frequency operation have to be considered. Those targets shall be achieved by considering the coaxial-cavity gyrotron technology and advanced technologies for key components (e.g. CVD diamond-disk Brewster angle window). Advanced simulation and test tools are complementing the research and developments. Gyrotron development is additionally supported by a significant investment into a new multi-megawatt long-pulse gyrotron test stand which is under final installation at KIT currently.

The effects of plasma shape on confinement and sawtooth stability are studied for positive and negative discharge triangularity and for different elongations with 1.5 MW centrally deposited ECH power

Basic wave-particle interaction dynamics from linear to chaotic regimes is experimentally studied on a frequency tunable gyrotron generating THz radiation in continuous mode (200W) at 263GHz which will be used for dynamic nuclear polarization nuclear magnetic resonance spectroscopy applications. In the studied system, the nonlinear dynamics associated to the waveparticle interaction is dominated by longitudinal mode competition of a given transverse TEm;p cavity-mode. This study covers a wide range of control parameter from gyro-traveling wave tube (gyro-TWT) to gyro-backward wave oscillator (gyro-BWO) like interactions for which extensive theoretical studies have been performed in the past on a simplified system. Besides the common route to chaos characterized by period doubling, other routes have been identified among which some are characterized by line-width frequency-broadening on the side-bands. The complex nonlinear dynamics is in good agreement with the theory and the experimental results are discussed on the basis of the prediction obtained with the nonlinear time-dependent selfconsistent codes TWANG and EURIDICE both based on a slow-time scale formulation of the self-consistent equations governing the wave-particle dynamics. VC

Within the range of plasma shapes and plasma currents investigated, the electron confinement time, tau_Ee, increases with safety factor, density and negative triangularity similar to the Ohmic heating case. There is little dependence of tau_Ee on the heating location provided power deposition occurs inside the q=1 surface; as power deposition moves out of the inversion surface, tau_Ee decreases. The power-induced energy confinement degradation exponent (tau_Ee~PaP) is as usual: alpha_P ~-0.5. As a general trend, central relaxations decrease in amplitude with increasing qa, P_EC, or negative delta, in a situation where the confinement time increases.

Magnetic fusion is one approach to generate thermonuclear fusion power in an environmental friendly way. The Electron Cyclotron Resonance Heating is considered as the major concept for startup, heating and control of the fusion plasma. Megawatt-class gyrotrons generate the required microwave power. This work focuses on advanced key components and technologies for a DEMO relevant 2 MW gyrotron. One major focus is on the development of advanced Magnetron Injection Guns. Another focus is on the red

Linac4 is the new negative hydrogen ion (H$^-$) linear accelerator of the European Organization for Nuclear Research (CERN). Its ion source operates on the principle of Radio-Frequency Inductively Coupled Plasma (RF-ICP) and it is required to provide 50~mA of H$^-$ beam in pulses of 600~$\mu$s with a repetition rate up to 2 Hz and within an RMS emittance of 0.25~$\pi$~mm~mrad in order to fullfil the requirements of the accelerator. This thesis is dedicated to the characterization of the hydrogen plasma in the Linac4 H$^-$ ion source. We have developed a Particle-In-Cell Monte Carlo Collision (PIC-MCC) code to simulate the RF-ICP heating mechanism and performed measurements to benchmark the fraction of the simulation outputs that can be experimentally accessed. The code solves self-consistently the interaction between the electromagnetic field generated by the RF coil and the resulting plasma response, including a kinetic description of charged and neutral species. A fully-implicit implementation allowed to simulate the high density regime of the Linac4 H$^-$ ion source, ensuring the energy conservation while maintaining the computational resources tractable. We studied the capacitive to inductive transition characteristic of the initial phase of the pulsed discharge. The simulation results were confirmed by time-resolved photometry measurements and allowed quantifying the effect of the hydrogen pressure and of the external magnetic cusp field on the transition dynamics. This provided insights into possible modifications to the magnetic cusp field configuration to maximize the power deposited to the plasma. The optimal ion source configuration maximizes the density of volume produced H$^-$, the flux of H$^0$ atoms onto the cesiated molybdenum plasma electrode surface at the origin of H$^-$ emission, and minimizes the electron density and energy in the beam formation region. We simulated the high-density regime (10$^{19}$ m$^{-3}$) representative of the nominal operation of the Linac4 ion source during beam extraction. We performed a parametric study of the RF current, hydrogen pressure and magnetic configuration (cusp and filter) to assess their impact on the plasma parameters. The simulation results allowed assessment of these parameters and provided guidelines for the optimization of the ion source operational and design parameters. The simulation results of electron density, electron energy and hydrogen dissociation degree showed excellent agreement with optical emission spectroscopy measurements both as a function of RF coil current and magnetic configuration. The outputs of these simulations provide crucial inputs to beam formation and extraction physics models. Dedicated PIC software packages are being developed that will eventually shed insight into essential beam parameters such as the intensity and emittance of the H$^-$ beam and the intensity of the co-extracted electrons.

The ring cyclotron of the Paul Scherrer Institute (PSI) accelerates an intense proton beam from 72MeV up to 590MeV. This happens in four cavities of very high quality factor, oscillating in the fundamental mode. The beam can excite parasitic oscillation modes (HOMs), because of its time structure. Measurements showed that their field can leak out into the vacuum chamber. Until now, there is no tool available to predict the potentially harmful effect of these HOMs onto the beam operation of the cyclotron. It is foreseeable that these effects might play a role if even higher beam currents have to be accelerated. This dissertation therefore deals with the numerical analysis and measurement of beam-cavity interactions. First calculations for a single cavity, interacting with a proton bunch were performed with MAFIA's eigenmode- (E3), time domain- (T3) and particle-in-cell (TS3) solvers. However, the structured grid and the limited computing performance of MAFIA make realistic simulations impossible. A simplified computation method is developed in this dissertation since a self-consistent simulation is impossible on today's computers: The parallel eigensolver Omega3P of the Stanford Linear Accelerator Center (SLAC) allowed us to calculate eigenmodes of the entire ring cyclotron for the first time ever. The rf fields are expanded onto a superposition of these modes and the excitation is calculated in frequency domain. Trajectories of the particles in the static magnetic field, superposed with the space charge fields and the beam excited HOMs, are then simulated. However, the quantitative accuracy of this model is still limited. On the one hand, because of the simplification in the geometry of the simulated rf structure, which otherwise would lead to a problem size going beyond the available computing resources. On the other hand, because it is not yet possible to simulate strongly absorbing boundaries more accurately. The simulation results confirm that up to proton beam currents of 2mA, corresponding to the routinely accelerated beam intensities, only a small deformation of the charge distribution appears. This thesis leads to a new simulation tool for further studies of intensity increases in high power cyclotrons.

As a result of extremely high upper critical fields B_c2, high temperature superconductors (HTS) have the potential to be used as high field insert coils in magnet systems where the background field is provided by low temperature superconductors (LTS) with the aim of application of such systems for high energy physics, nuclear magnetic resonance and energy storage. Among the superconductors discovered in late 1980s, one that is widely studied is YBCO coated conductor (CC). Despite the advanced 2G conductor technology in recent years allowing for manufacturing of long lengths of YBCO CC, up to 1 km, the crucial issue in practical application of the hybrid systems with YBCO insert coil remain the quench detection and protection of the insert. Unlike in LTS magnets, the quench propagation in HTS is significantly slower, which makes quench detection and consecutively protection, very challenging. Following a new approach for quench protection a non-insulated HTS insert coil was manufactured at the CRPP. In case of quench, the current is expected to by-pass the normal zone and prevent the coil from a permanent damage. The idea of non-insulated free-standing coils has been already studied and brought promising results in terms of self-protection of such coils. However, in a hybrid system of LTS-HTS, the quench behaviour needs careful evaluation that includes magneto-thermo-electrical study. The objective of this work is to study and discuss the applicability of HTS insert coils (insulated and non-insulated) by addressing the issue of their quench detection and protection schemes. The starting point for the analysis is studying the existing design according to CRPP specifications taking into account various operating modes together with detection and protection schemes. Finally, general guidelines for the design of a successful LTS-HTS magnet system will be discussed.

The quality factor and oscillating mode of a gyrotron cavity are essential parameters to consider when trying to obtain a high power (>500 kW), high efficiency (∼50%) microwave source, which oscillates in a stable manner in the principal mode of the cavity. The study and development of an 8 GHz gyrotron whose resonant cavity is formed by a cylindrical waveguide of slowly varying radius, is undertaken. The study is principally concerned with the phenomena associated with the low quality factor of the TE°011 mode of the cavity. The power at optimal efficiency of a given resonant mode (TE°011 ) is increased when the diffractive quality factor (Qdiff) of the mode is decreased. Experiments performed with the gyrotron delivering power into a matched load show that during the process of decreasing Qdiff, the chosen mode of oscillation is perturbed by the appearance of oscillations in parasitic modes of the cavity; composed of higher order longitudinal modes (TE°O1q'q=2.3), and a propagating mode TE°21' gyro BWO), corresponding to an absolute instability of the system. These modes limit the power and efficiency of the principal mode by either oscillating simultaneously with, or in place of, the principal mode. When the gyro BWO mode oscillates, even at low power (several kilowatts), the efficiency of the principal mode of a cavity with Qdiff=160 is limited to less than 25%. This parasitic oscillation can be eliminated by the addition of a gradient in the magnetic field over the length of the interaction region. The maximum efficiency of the principal mode is then greater thon 40%; however, the power delivered remains lower than that of the cavity with Qdiff=225. The study of the influence of power reflections on the oscillation characteristics of the gyrotron reveals that the lowering of the quality factor of the cavity increases the sensitivity of the gyrotron to power reflections due to mismatched loads. A gyrotron with a quality factor which is too low can only deliver its maximum power to a matched load. In the presence of power reflections, the mode of oscillation changes, or the pulse is interrupted by an arc in the interior of the gyrotron. Even though a maximum power of 310 kW at an efficiency of 35% is obtained by the cavity with Qdiff=225, the sensitivity of this gyrotron to power reflections could limit the possibility of its use as a source for current generation or heating in a plasma, for example. These results show that there is a lower limit to the quality factor of a gyrotron cavity. This limit is situated between 410 and 225 for the TE°011 mode at 8 GHz. Below this limit the power in the mode decreases, the oscillation is perturbed or eliminated by the simultaneous oscillation of parasitic modes, and the sensitivity to power reflections could render the tube unusable for practical applications. A determination of the average velocity ratio α of the electrons in the electron beam (α=/) is made by measuring the oscillation frequency of the gyro BWO mode. The results obtained using a non-linear simulation of the interaction between the gyro BWO and the electron beam show that the value of α at the entrance of the cavity corresponds to the α-value which the electron gun is supposed to produce. It seems, however, to indicate the presence of a larger velocity dispersion than expected. This original method of measuring the α of an electron beam can be applied to any source in which a gyro BWO mode is observed, or directly as a method of characterising the beam generated by a given electron gun.

One of the biggest goals for the international thermonuclear experimental reactor (ITER) is the steady state operation. For this reason all its coils will be superconducting and Nb3Sn material is used for both the toroidal field (TF) and the central solenoid (CS) coils. The primary current carrying unit in a coil is the cable, which contains hundreds of superconducting strands. The ITER coils will be made using a kind of cable called cable in conduit conductor (CICC). Both CS and TF magnets have to withstand several electromagnetic (EM) cycles in their lifetime. This has been discovered to cause a severe reduction of the magnet performances. Nb3Sn is a brittle material, therefore it is very strain sensitive. The knowledge of the strain state in a CICC allows the explanation of the performances of the cable itself, providing also information regarding the permanent damaging of the Nb3Sn structure. The measurement of the Nb3Sn strain state in a CICC is a very challenging task because it is not possible to make it directly. An indirect technique that allows the measurement of the thermal strain through the measurement of the magnetic susceptibility vs. temperature (χ(T )) for a CICC has been developed. A CICC is a complex system and as such it behaves in a way which is not directly predictable studying the single components. The elastic deformation of the superconducting layers produces reversible phenomena which interact with irreversible phenomena due to Nb3Sn fractures and to plastic deformations of the matrix surrounding the superconducting layers. Reversible and irreversible phenomena occurs together and are deeply shuffled such that it is not straightforward the explanation of the cable performance. The magnetic susceptibility measurements, via an appropriate data analysis, allow to quantify and to separate the thermal strain from the jumble of parameters necessary to describe the cable performance. Several TF and CS samples have been measured with this technique in the SULTAN test facility. Each cable behaves differently from the others but at least it is possible to identify common features. For some samples, the evolution of the performance with the EM cycles is dominated by reversible rearrangements of the strand position in the cable and consequently by the change of the strain distribution during the EM load, for others it is dominated by breakages and irreversible phenomena.