The design of fusion devices is typically based on computationally expensive simulations. This can be alleviated using high aspect ratio models that employ a reduced number of free parameters, especially in the case of stellarator optimization where non-axisymmetric magnetic fields with a large parameter space are optimized to satisfy certain performance criteria. However, optimization is still required to find configurations with properties such as low elongation, high rotational transform, finite plasma beta, and good fast particle confinement. In this work, we train a machine learning model to construct configurations with favorable confinement properties by finding a solution to the inverse design problem, that is, obtaining a set of model input parameters for given desired properties. Since the solution of the inverse problem is non-unique, a probabilistic approach, based on mixture density networks, is used. It is shown that optimized configurations can be generated reliably using this method.
William & Mary（威廉与玛丽学院）；Northeastern University（东北大学）；Augusta University（奥古斯塔大学）
Deep Neural Networks (DNNs) have emerged as the core enabler of many major applications on mobile devices. To achieve high accuracy, DNN models have become increasingly deep with hundreds or even thousands of operator layers, leading to high memory and computational requirements for inference. Operator fusion (or kernel/layer fusion) is key optimization in many state-of-the-art DNN execution frameworks, such as TensorFlow, TVM, and MNN. However, these frameworks usually adopt fusion approaches based on certain patterns that are too restrictive to cover the diversity of operators and layer conn…
A design principle is suggested to overcome obstacles that prevent positive net energy output in nuclear fusion devices based on electrostatically accelerated ions. Since Coulomb scattering cross-section dwarfs that of nuclear fusion, the focus is on re-capturing energy of elastically scattered ions before the energy is lost to heat. Device configuration to achieve efficient energy re-capturing is proposed and a favorable estimate of net energy gain is obtained.
In a variety of magnetized plasma geometries, it has long been known that highly charged impurities tend to accumulate in regions of higher density. Here, we examine how this "collisional pinch" changes in the presence of additional forces, such as might be found in systems with gravity, fast rotation, or non-negligible space charge. In the case of a rotating, cylindrical plasma, we describe a regime in which the radially outermost ion species is intermediate in both mass and charge. We discuss implications for fusion devices and plasma mass filters.
Theory for a gridded inertial electrostatic confinement (IEC) fusion system is presented that shows a net energy gain is possible if the grid is magnetically shielded from ion impact. A simplified grid geometry is studied, consisting of two negatively-biased coaxial current-carrying rings, oriented such that their opposing magnetic fields produce a spindle cusp. Our analysis indicates that better than break-even performance is possible even in a deuterium-deuterium system at bench-top scales. The proposed device has the unusual property that it can avoid both the cusp losses of traditional mag…
The lower hybrid wave trapping in a tokamak edge transport barrier is predicted. This effect makes possible excitation of the low-power-threshold absolute parametric decay instability leading to side-scattering of the ordinary microwave pump in the electron cyclotron resonance heating experiments. The instability can result in broadening of the power deposition profile both in present day machines and in ITER.
The turbulence in magnetically confined fusion plasmas has important and non-trivial effects on the quality of the energy confinement. These effects are hard to make a quantitative assessment of analytically. The problem investigated in this article is the transport of energy and particles, in particular impurities, in a Tokamak plasma. Impurities from the walls of the plasma vessel cause energy losses if they reach the plasma core. It is therefore important to understand the transport mechanisms to prevent impurity accumulation and minimize losses. This is an area of research where turbulence…
A fusion reactor requires plasma pre-heating before the rate of deuterium-tritium fusion reactions becomes significant. In ITER, radiofrequency (RF) heating of 3He ions, additionally puffed into the plasma, is one of the main options considered for increasing bulk ion temperature during the ramp-up phase of the pulse. In this paper, we propose an alternative scenario for bulk ion heating with RF waves, which requires no extra 3He puff and profits from the presence of intrinsic Beryllium impurities in the plasma. The discussed method to heat Be impurities in D-T plasmas is shown to provide an e…
Evidences for and the models of self-similar skeletal structures in fusion devices, severe weather phenomena and space
本文简要回顾了（i）在10-5 cm - 1023 cm范围内观察到的一种骨骼形式（即小管和侧手翻及其最简单的组合）的自相似结构（称为通用骨骼结构（USS））的证据。在高电流放电在各种融合设备，恶劣的天气现象，和空间，（ii）解释骨架结构的现象，包括一个分形凝聚态（FCM）的假设模型，从纳米管灰尘组装，及（iii）可能的作用FCM，这可能是负责的USS现象，在龙卷风，球形闪电，和水龙卷。
The conventional inertial electrostatic confinement fusion (IECF) operation is based on the application of high negative voltage to the central grid which results in the production of neutrons due to the fusion of lighter ions. The neutron has enormous applications in diversified fields. Apart from the neutrons, it can also be used as an application based x-ray source by altering the polarity of the central grid. In this work, the electron dynamics during the positive polarity of the central grid have been studied using an object-oriented particle-in-cell code (XOOPIC). The trapped electron de…
The multi-scale, mutli-physics nature of fusion plasmas makes predicting plasma events challenging. Recent advances in deep convolutional neural network architectures (CNN) utilizing dilated convolutions enable accurate predictions on sequences which have long-range, multi-scale characteristics, such as the time-series generated by diagnostic instruments observing fusion plasmas. Here we apply this neural network architecture to the popular problem of disruption prediction in fusion tokamaks, utilizing raw data from a single diagnostic, the Electron Cyclotron Emission imaging (ECEi) diagnostic…
To guarantee the availability and reliability of data source in Magnetic Confinement Fusion (MCF) devices, incorrect diagnostic data, which cannot reflect real physical properties of measured objects, should be sorted out before further analysis and study. Traditional data sorting cannot meet the growing demand of MCF research because of the low-efficiency, time-delay, and lack of objective criteria. In this paper, a Time-Domain Global Similarity (TDGS) method based on machine learning technologies is proposed for the automatic data cleaning of MCF devices. Traditional data sorting aims to the…
In order to improve the presently used ad hoc flux limiter treatment of parallel heat flux transport in edge plasma fluid codes we consider here a generalized version of the Fourier law implementing a non-local kernel for the heat flux computation. The Bohm boundary condition at the wall is recovered introducing a volumetric loss term representing the contribution of suprathermal particles to the energy out flux. As expected, this contribution is negligible in the strongly collisional regime while it becomes more and more dominant for marginally and low collisional regimes. In the second part …
Edge turbulent structures are commonly observed in fusion devices and are generally believed to be responsible for confinement degradation. Among their origin Drift-Alfven turbulence is one of the most commonly suggested. Drift-Alfven paradigm allows the existence of localized vortex-like structures observed also in various systems. Here we present the evidence of the presence of drift-Alfvén vortices in the edge region of RFX-Mod RFP device, showing how these structures are responsible for electromagnetic turbulence at the edge and its intermittent nature.
École Polytechnique Fédérale de Lausanne（洛桑联邦理工学院）
The analysis of turbulence in plasmas is fundamental in fusion research. Despite extensive progress in theoretical modeling in the past 15 years, we still lack a complete and consistent understanding of turbulence in magnetic confinement devices, such as tokamaks. Experimental studies are challenging due to the diverse processes that drive the high-speed dynamics of turbulent phenomena. This work presents a novel application of motion tracking to identify and track turbulent filaments in fusion plasmas, called blobs, in a high-frequency video obtained from Gas Puff Imaging diagnostics. We comp…
The Tokamak device is the most promising candidate for producing sustainable electric power by nuclear fusion. It is a torus-shaped device that confines plasma by a strong magnetic field. The development, design and control of the design has been an important area of research, and a significant target is to effectively confine the extremely hot plasma inside its hollow torus-shaped body without touching its boundary for a prolonged period of time. In an attempt to control a Tokamak device, this paper investigates an optimal control problem for an incompressible, viscous, electrically conductin…
In contrast to the broad stream of sustainable developments on fusion energy, new aspects are developed now by applying ultra-short, ultra-powerful laser pulses in a plasma-block ignition scheme by avoiding the well known difficulties of thermal-pressure instabilities and losses through using electro-dynamic non-thermal energy conversion. A further advantage is given by the new 10 kilo-Tesla magnetic fields for fusion of uncompressed proton-boron fuel which avoids problems of dangerous nuclear radiation.
Nuclear fusion is the process that powers the sun, and it is one of the best hopes to achieve a virtually unlimited energy source for the future of humanity. However, reproducing sustainable nuclear fusion reactions here on Earth is a tremendous scientific and technical challenge. Special devices -- called tokamaks -- have been built around the world, with JET (Joint European Torus, in the UK) being the largest tokamak currently in operation. Such devices confine matter and heat it up to extremely high temperatures, creating a plasma where fusion reactions begin to occur. JET has over one hund…
In the era of big science, countries allocate big research and development budgets to large scientific facilities that boost collaboration and research capability. A nuclear fusion device called the "tokamak" is a source of great interest for many countries because it ideally generates sustainable energy expected to solve the energy crisis in the future. Here, to explore the scientific effects of tokamaks, we map a country's research capability in nuclear fusion research with normalized revealed comparative advantage on five topical clusters -- material, plasma, device, diagnostics, and simula…
Recently-proposed tokamak concepts use magnetic fields up to 12 T, far higher than in conventional devices, to reduce size and cost. Theoretical and computational study of trends in plasma behavior with increasing field strength is critical to such proposed devices. This paper considers trends in Alfvén eigenmode (AE) stability. Energetic particles, including alphas from D-T fusion, can destabilize AEs, possibly causing loss of alpha heat and damage to the device. AEs are sensitive to device magnetic field via the field dependence of resonances, alpha particle beta, and alpha orbit width. We d…
The first combined experimental and theoretical studies of dust remobilization by plasma forces are reported. The main theoretical aspects of remobilization in fusion devices under steady state conditions are analyzed. In particular, the dominant role of adhesive forces is highlighted and generic remobilization conditions - direct lift-up, sliding, rolling - are formulated. A novel experimental technique is proposed, based on controlled adhesion of dust grains on tungsten samples combined with detailed mapping of the dust deposition profile prior and post plasma exposure. Proof-of-principle ex…
In inertial confinement fusion, the scientific issues include the generation and transport of driver energy, the pellet design, the uniform target implosion physics, the realistic nuclear fusion reactor design, etc. In this paper, we present a pellet injection into a power reactor in heavy ion inertial fusion. We employ a magnetic correction method to reduce the pellet alignment error in heavy ion inertial fusion reactor chamber, including the gravity, the reactor gas drag force and the injection errors. We found that the magnetic correction device proposed in this paper is effective to constr…
In magnetically-confined plasmas of tokamaks, neutral deuterium/hydrogen (D/H) atoms play a role in energy, momentum, and particle balance, as well as the stabilization of plasma turbulence. One key important fusion performance parameter is the pedestal density. Understanding the pedestal density formation is critical for the development of predictive model of future fusion devices. Typically, measurements of the neutrals are obtained using optical emission spectroscopy of the Lyman alpha lines, which is a line-integrated measurement. The plasma in tokamaks is characterized by a high density o…
Thermonuclear fusion reactivity is a pivotal quantity in the studies pertaining to fusion energy production, fusion ignition and energy break-even analysis in both inertially and magnetically confined systems. Although nuclear fusion reactivity and thereafter the power density of a magnetic confinement fusion reactor and the fulfillment of the ignition criterion are quantitatively determined by assuming the ion speed distribution to be Maxwellian, a significant population of suprathermal ions,with energy greater than the quasi-Maxwellian background plasma temperature, is generated by the fusio…
Degradation of first wall materials due to plasma disturbances severely limit both the lifetime and longevity of fusion reactors. Among the various kinds of disturbances, type I edge localized modes (ELMs) in particular present significant design challenges due to their expected heat loading and relative frequency in next step fusion reactors. Plasma gun devices have been used extensively to replicate ELM conditions in the laboratory, however feature higher density, lower temperatures, and thus higher flow collisionality than those expected in fusion conditions. This work presents experimental…
A Self-Organized-Criticality model consistent with statistical properties of edge turbulence in a fusion plasma
研究了最近的一个自组织临界性（SOC）模型产生的间歇信号的统计特性。一个成功的比较与先前公布的结果的等效量测量的静电湍流在聚变等离子体的边缘。这一结果重新确立了SOC作为磁聚变装置中传输的潜在范例，克服了早期工作中指出的缺点[E。Spada等人，Phys.Rev.Lett. 86，3032（2001）; V. Antoni等人，Phys.Rev.Lett. 87，045001（2001）]。
Institute for Plasmas and Nuclear Fusion（等离子体与核聚变研究所）；IST, University of Lisbon（里斯本大学高等技术学院）；Culham Science Centre（卡尔姆科学中心）
The use of deep learning is facilitating a wide range of data processing tasks in many areas. The analysis of fusion data is no exception, since there is a need to process large amounts of data collected from the diagnostic systems attached to a fusion device. Fusion data involves images and time series, and are a natural candidate for the use of convolutional and recurrent neural networks. In this work, we describe how CNNs can be used to reconstruct the plasma radiation profile, and we discuss the potential of using RNNs for disruption prediction based on the same input data. Both approaches…
