• Themes
• Newsmakers

# Fusion

25.06.2022
02:33 A Garage-Sized Reactor Could Provide Limitless Energy With Magnet-Free Technology (Slashdot)

SlashdotA Garage-Sized Reactor Could Provide Limitless Energy With Magnet-Free Technology - An anonymous reader quotes a report from Interesting Engineering: Seattle-based Zap Energy is using a lesser-known approach to nuclear fusion to build modular, garage-sized reactors. They are cheaper and dont require the large, incredibly powerful magnets used in traditional fusion experiments. Ultimately, they may also provide a quicker route to ...

24.06.2022
10:03 Probabilistic load forecasting for the low voltage network: forecast fusion and daily peaks. (arXiv:2206.11745v1 [stat.AP])

Short-term forecasts of energy consumption are invaluable for operation of energy systems, including low voltage electricity networks. However, network loads are challenging to predict when highly desegregated to small numbers of customers, which may be dominated by individual behaviours rather than the smooth profiles associated with aggregate consumption. Furthermore, distribution networks are challenged almost entirely by peak loads, and tasks such as scheduling storage and/or demand flexibility maybe be driven by predicted peak demand, a feature that is often poorly characterised by general-purpose forecasting methods. Here we propose an approach to predict the timing and level of daily peak demand, and a data fusion procedure for combining conventional and peak forecasts to produce a general-purpose probabilistic forecast with improved performance during peaks. The proposed approach is demonstrated using real smart meter data and a hypothetical low voltage network hierarchy

10:03 Fusion of Model-free Reinforcement Learning with Microgrid Control: Review and Insight. (arXiv:2206.11398v1 [eess.SY])

Challenges and opportunities coexist in microgrids as a result of emerging large-scale distributed energy resources (DERs) and advanced control techniques. In this paper, a comprehensive review of microgrid control is presented with its fusion of model-free reinforcement learning (MFRL). A high-level research map of microgrid control is developed from six distinct perspectives, followed by bottom-level modularized control blocks illustrating the configurations of grid-following (GFL) and grid-forming (GFM) inverters. Then, mainstream MFRL algorithms are introduced with an explanation of how MFRL can be integrated into the existing control framework. Next, the application guideline of MFRL is summarized with a discussion of three fusing approaches, i.e., model identification and parameter tuning, supplementary signal generation, and controller substitution, with the existing control framework. Finally, the fundamental challenges associated with adopting MFRL in microgrid control and

23.06.2022
21:32 Fusion power may run out of fuel before it even gets started

Experts fear giant ITER reactor will worsen looming shortage of tritium

05:43 Theoretical and experimental study on Noise Equivalent Power of X-ray semiconductor ultra-fast response material based on the rad-optic effect. (arXiv:2206.10890v1 [physics.ins-det])

Semiconductor material based on the rad-optic effect enables ultra-fast detection of X-rays and plays an important role in fusion diagnostics. Obtaining the accurate noise equivalent power (NEP) of the semiconductor ultrafast response material is the key to detecting X-rays. In this paper, the refractive index change mechanism of the semiconductor under X-ray irradiation was analyzed, and the quantitative relationship between the diffraction efficiency and the X-ray photon energy was established through the LT-AlGaAs diffraction imaging experiments. The impulse responses of LT-AlGaAs under 1 KeV-10 KeV X-ray radiation were calculated, revealing the variation of NEP density with radiated photon energy. In the case of bombarding the Al target to generate 1.5 KeV X-rays, the imaging experiments of LT-AlGaAs were performed. The diffraction image of LT-AlGaAs has a linear relationship with the radiation intensity, and the NEP density of LT-AlGaAs reaches 4.80*105W/cm2. This study has

22.06.2022
22:33 More funding for fusion: Seattle startup lands $160M and reveals technology breakthrough (Lisa Stiffler/GeekWire) Lisa Stiffler / GeekWireMore funding for fusion: Seattle startup lands$160M and reveals technology breakthrough - The news: Seattle-area startup Zap Energy is edging closer to the promise of fusion power. The company Wednesday announced two milestones: its newest prototype device has created plasma, a superheated gas needed to generate fusion, and it raised $160 million in new funding with support from Bill Gates Breakthrough Energy ... 13:01 Fusion Energy Advance Is Hailed by a Seattle Start-Up Zap Energy said its experimental reactor core was ready for a milestone test. Skeptics routinely question whether the technology is currently possible. 12:02 Fusion Energy Advance Is Hailed by a Seattle Start-Up Zap Energy said its experimental reactor core was ready for a milestone test. Skeptics routinely question whether the technology is currently possible. 12:02 Fusion Energy Advance Is Hailed by a Seattle Start-Up Zap Energy said its experimental reactor core was ready for a milestone test. Skeptics routinely question whether the technology is currently possible. 08:23 Data-driven model for divertor plasma detachment prediction. (arXiv:2206.09964v1 [physics.plasm-ph]) We present a fast and an accurate data-driven surrogate model for divertor plasma detachment prediction leveraging the latent space concept in machine learning research. Our approach involves constructing and training two neural networks. An autoencoder that finds a proper latent space representation (LSR) of plasma state by compressing the multi-modal diagnostic measurements, and a forward model using multi-layer perception (MLP) that projects a set of plasma control parameters to its corresponding LSR. By combining the forward model and the decoder network from autoencoder, this new data-driven surrogate model is able to predict a consistent set of diagnostic measurements based on a few plasma control parameters. In order to ensure that the crucial detachment physics is correctly captured, highly efficient 1D UEDGE model is used to generate training and validation data in this study. Benchmark between the data-driven surrogate model and UEDGE simulations shows that our surrogate 08:23 A Machine Learning Data Fusion Model for Soil Moisture Retrieval. (arXiv:2206.09649v1 [physics.ao-ph]) We develop a deep learning based convolutional-regression model that estimates the volumetric soil moisture content in the top ~5 cm of soil. Input predictors include Sentinel-1 (active radar), Sentinel-2 (optical imagery), and SMAP (passive radar) as well as geophysical variables from SoilGrids and modelled soil moisture fields from GLDAS. The model was trained and evaluated on data from ~1300 in-situ sensors globally over the period 2015 - 2021 and obtained an average per-sensor correlation of 0.727 and ubRMSE of 0.054, and can be used to produce a soil moisture map at a nominal 320m resolution. These results are benchmarked against 13 other soil moisture works at different locations, and an ablation study was used to identify important predictors. 20.06.2022 06:43 Experimental scaling of the scrape-off layer particle flux width by outboard divertor Langmuir probes with favorable Bt configuration on EAST. (arXiv:2206.08643v1 [physics.plasm-ph]) The scrape-off layer (SOL) power width {\lambda}q is important for predicting the heat load on divertor targets for future magnetically confined devices. However, there are still some inconsistencies between the experimental and simulation results for {\lambda}q scaling. This paper extends the previous SOL particle flux width (\lambda_{js}) scaling [Liu et al 2019 Plasma Phys. Control. Fusion 61 045001] and provides more experimental evidence to support the {\lambda}q study. A systematic method has been developed to calibrate the upper outer (UO) divertor Langmuir probes (Div-LPs) with the Ohmic discharges to reduce the measurement uncertainty of \lambda_{js}. For the discharges with the favorable Bt and upper single null configurations in the 2019 EAST experiment campaign, about 260 discharges have been filtered out. The calibrated js data measured by the UO Div-LPs have been statistically analyzed. Three H-mode, L-mode, and Ohmic scaling databases have been constructed and are used 06:43 Development of an interactive code for quick data analyses between STOR-M tokamak experimental plasma discharges. (arXiv:2206.08458v1 [physics.plasm-ph]) Saskatchewan Torus-Modified (STOR-M) is a small tokamak, well known for various fusion related basic experimental studies such as edge turbulent heating, different instabilities, AC (alternating current) tokamak operation, Ohmic H-mode triggering by the electrode biasing, fueling and momentum injection by Compact Torus (CT) injection, and effects of Resonance Magnetic Perturbations (RMP), among others. Some of those experiments require real time visualization of magnetic surface reconstructions either through EFIT or quick analyses and visualization of experimental data during experiments. Recently experimental studies of Geodesic Acoustic Mode (GAM) and zonal flows had been performed in STOR-M tokamak. The GAM experiments strongly require collection of fluctuations data from different Langmuir probes installed at different poloidal locations, but on the same magnetic surfaces. This is need of the adjustment of radial locations between discharges. It is therefore important to analyze 06:43 Direct construction of stellarator-symmetric quasi-isodynamic magnetic configurations. (arXiv:2206.08417v1 [physics.plasm-ph]) We develop the formalism of the first order near-axis expansion of the MHD equilibrium equations described in Garren & Boozer (1991), Plunk et al. (2019) and Plunk et al. (2021), for the case of a quasi-isodynamic, N-field period, stellarator symmetric, single-well magnetic field equilibrium. The importance of the magnetic axis shape is investigated, and we conclude that control of the curvature and torsion is crucial to obtain omnigenous configurations with finite aspect ratio and low effective ripple, especially for a higher number of field periods. For this reason a method is derived to construct classes of axis shapes with favourable curvature and torsion. Solutions are presented, including a three-field-period configuration constructed at an aspect ratio of A=20, with a maximum elongation of e=3.2 and an effective ripple under 1%, which demonstrates that high elongation is not a necessary feature of QI stellarators. 06:43 An primary data management for Fusion AI study. (arXiv:2206.08414v1 [physics.plasm-ph]) A data management for fusion AI research based on MongoDB and Hierarchical Data Format version 5 (HDF5) has been developed and implemented on EAST tokamak. Currently, the entire data management has more than 3000 channels of data. The system can provide more than 100 highly reliable concurrent access processes simultaneously. The system includes error correction, MDSplus original data conversion, and high-performance sequence data output. Moreover, some useful functions are implemented to accelerate fusion AI model training, such as bucketing, the concatenating buffer, and distributed sequence generation. This data management system is more suitable for fusion AI model R&D than MDSplus, but it can not replace the MDSplus database. The MDSplus database is still the backend for EAST tokamak data acquisition and storage. 17.06.2022 07:52 Turbulent Transport in Tokamak-plasmas: A Thermodynamic Approach. (arXiv:2206.08310v1 [physics.plasm-ph]) In a previous work we provided the explicit form of the nonlinear PDEs, subjected to the appropriate boundary conditions, which have to be satisfied by transport coefficients for systems out of Onsager's region. Since the proposed PDEs are obtained without neglecting any term present in the balance equations (i.e., the mass, momentum, and energy balance equations), we propose them as a good candidate for describing transport in thermodynamic systems also in turbulent regime. As a special case, we derive the nonlinear PDEs for transport coefficients when the thermodynamic system is subjected to two thermodynamic forces. In this case, the obtained PDE is, in thermodynamical field theory (TFT), analogous to Liouville's equation in Riemannian (or pseudo-Riemannian) geometry. The validity of our model is tested by analysing a concrete example where Onsager's relations manifestly disagree with experience: transport in Tokamak-plasmas. More specifically, we compute the electron mass and 07:52 Hardness prediction of age-hardening aluminum alloy based on ensemble learning. (arXiv:2206.08011v1 [cond-mat.mtrl-sci]) With the rapid development of artificial intelligence, the combination of material database and machine learning has driven the progress of material informatics. Because aluminum alloy is widely used in many fields, so it is significant to predict the properties of aluminum alloy. In this thesis, the data of Al-Cu-Mg-X (X: Zn, Zr, etc.) alloy are used to input the composition, aging conditions (time and temperature) and predict its hardness. An ensemble learning solution based on automatic machine learning and an attention mechanism introduced into the secondary learner of deep neural network are proposed respectively. The experimental results show that selecting the correct secondary learner can further improve the prediction accuracy of the model. This manuscript introduces the attention mechanism to improve the secondary learner based on deep neural network, and obtains a fusion model with better performance. The R-Square of the best model is 0.9697 and the MAE is 3.4518HV. 07:52 Challenges and Opportunities in Deep Reinforcement Learning with Graph Neural Networks: A Comprehensive review of Algorithms and Applications. (arXiv:2206.07922v1 [cs.LG]) Deep reinforcement learning (DRL) has empowered a variety of artificial intelligence fields, including pattern recognition, robotics, recommendation-systems, and gaming. Similarly, graph neural networks (GNN) have also demonstrated their superior performance in supervised learning for graph-structured data. In recent times, the fusion of GNN with DRL for graph-structured environments has attracted a lot of attention. This paper provides a comprehensive review of these hybrid works. These works can be classified into two categories: (1) algorithmic enhancement, where DRL and GNN complement each other for better utility; (2) application-specific enhancement, where DRL and GNN support each other. This fusion effectively addresses various complex problems in engineering and life sciences. Based on the review, we further analyze the applicability and benefits of fusing these two domains, especially in terms of increasing generalizability and reducing computational complexity. Finally, the 16.06.2022 10:32 Accurate Emotion Strength Assessment for Seen and Unseen Speech Based on Data-Driven Deep Learning. (arXiv:2206.07229v1 [cs.SD]) Emotion classification of speech and assessment of the emotion strength are required in applications such as emotional text-to-speech and voice conversion. The emotion attribute ranking function based on Support Vector Machine (SVM) was proposed to predict emotion strength for emotional speech corpus. However, the trained ranking function doesn't generalize to new domains, which limits the scope of applications, especially for out-of-domain or unseen speech. In this paper, we propose a data-driven deep learning model, i.e. StrengthNet, to improve the generalization of emotion strength assessment for seen and unseen speech. This is achieved by the fusion of emotional data from various domains. We follow a multi-task learning network architecture that includes an acoustic encoder, a strength predictor, and an auxiliary emotion predictor. Experiments show that the predicted emotion strength of the proposed StrengthNet is highly correlated with ground truth scores for both seen and unseen 15.06.2022 06:03 Locked mode disruptions in DIII-D and application to ITER. (arXiv:2206.06773v1 [physics.plasm-ph]) Disruptions are rapid loss of plasma confinement in tokamaks, which could damage large tokamaks like ITER. Recent work identified the thermal quench in JET locked mode disruptions with a resistive wall tearing mode. New research finds a similar instability in a DIII-D locked mode shot. The instability is studied with simulations, theory, and comparison to experimental data. Linear theory is extended to include resistive wall modes with a rational surface in the plasma. Linear simulations show the mode is stable for an ideal wall, and unstable with a resistive wall. Nonlinear simulations show that the mode grows to large amplitude, causing a thermal quench. The mode onset occurs when the radius of the q =2 rational surface is sufficiently close to the plasma edge. These results are important for ITER, greatly mitigating the effects of disruptions. 06:03 Constraints on stellarator divertors from Hamiltonian mechanics. (arXiv:2206.06368v1 [physics.plasm-ph]) The design of any large stellarator requires a plan for the removal of the particles and heat that are exhausted across the plasma edge. This is called the divertor problem, for the particle exhaust must be diverted into pumping chambers. Although the physics of diverted plasmas has many subtleties, the magnetic field configuration between the plasma edge and the surrounding chamber walls is the foundation upon which divertor design is based. The properties of this magnetic configuration has both practical constraints and mathematical constraints from magnetic field lines obeying a 1~1/2 degree of freedom Hamiltonian. Practical constraints from plasma physics will be discussed to the extent needed to integrate with the constraints from from Hamiltonian mechanics for the conceptual design of stellarator divertors. 06:03 Does a Technique for Building Multimodal Representation Matter? -- Comparative Analysis. (arXiv:2206.06367v1 [cs.LG]) Creating a meaningful representation by fusing single modalities (e.g., text, images, or audio) is the core concept of multimodal learning. Although several techniques for building multimodal representations have been proven successful, they have not been compared yet. Therefore it has been ambiguous which technique can be expected to yield the best results in a given scenario and what factors should be considered while choosing such a technique. This paper explores the most common techniques for building multimodal data representations -- the late fusion, the early fusion, and the sketch, and compares them in classification tasks. Experiments are conducted on three datasets: Amazon Reviews, MovieLens25M, and MovieLens1M datasets. In general, our results confirm that multimodal representations are able to boost the performance of unimodal models from 0.919 to 0.969 of accuracy on Amazon Reviews and 0.907 to 0.918 of AUC on MovieLens25M. However, experiments on both MovieLens datasets 14.06.2022 08:42 COLD Fusion: Calibrated and Ordinal Latent Distribution Fusion for Uncertainty-Aware Multimodal Emotion Recognition. (arXiv:2206.05833v1 [cs.CV]) Automatically recognising apparent emotions from face and voice is hard, in part because of various sources of uncertainty, including in the input data and the labels used in a machine learning framework. This paper introduces an uncertainty-aware audiovisual fusion approach that quantifies modality-wise uncertainty towards emotion prediction. To this end, we propose a novel fusion framework in which we first learn latent distributions over audiovisual temporal context vectors separately, and then constrain the variance vectors of unimodal latent distributions so that they represent the amount of information each modality provides w.r.t. emotion recognition. In particular, we impose Calibration and Ordinal Ranking constraints on the variance vectors of audiovisual latent distributions. When well-calibrated, modality-wise uncertainty scores indicate how much their corresponding predictions may differ from the ground truth labels. Well-ranked uncertainty scores allow the ordinal ranking 08:42 Fusing Feature Engineering and Deep Learning: A Case Study for Malware Classification. (arXiv:2206.05735v1 [cs.CR]) Machine learning has become an appealing signature-less approach to detect and classify malware because of its ability to generalize to never-before-seen samples and to handle large volumes of data. While traditional feature-based approaches rely on the manual design of hand-crafted features based on experts knowledge of the domain, deep learning approaches replace the manual feature engineering process by an underlying system, typically consisting of a neural network with multiple layers, that perform both feature learning and classification altogether. However, the combination of both approaches could substantially enhance detection systems. In this paper we present an hybrid approach to address the task of malware classification by fusing multiple types of features defined by experts and features learned through deep learning from raw data. In particular, our approach relies on deep learning to extract N-gram like features from the assembly language instructions and the bytes of 13.06.2022 17:53 Helium pre-exposure inhibits hydrogen isotope permeation in wall materials A research team from the Hefei Institutes of Physical Science (HFIPS) of the Chinese Academy of Sciences (CAS) has recently revealed that helium exposure could inhibit the penetration of hydrogen isotopes in wall materials. Their results were published in Nuclear Fusion. 04:12 Plane strain optimization of conductor and structure grading in the inner leg of a Tokamak toroidal field coil. (arXiv:2206.04712v1 [physics.plasm-ph]) I present the results of the analytic plane strain optimization of the structure and conductor grading in the inner leg of a Tokamak toroidal field coil. The coil is assumed to be made of regions of soft conductor inside a stiff grid of conduit. Calculus of variations is used to determine the optimal profile of this structure. The optimal solution is found to be two-layered. An outer layer is bucked only, not wedged, and the structure fraction is graded so that all structure is at a uniform stress. An inner layer is an advanced bucking cylinder, similar to a Florida Bitter plate, whose radial stiffness is tuned so that its azimuthal stress is uniform. Such an advanced bucking cylinder would require advanced manufacturing to fabricate. These results should be seen an upper limits rather than achievable performance. Concepts that arise from this optimization, such as selectively softening layers of the bucking cylinder by drilling or cutting, may be of use to real designs in a 04:12 Computational thermal multi-phase flow for metal additive manufacturing. (arXiv:2206.04799v1 [cs.CE]) Thermal multi-phase flow simulations are indispensable to understanding the multi-scale and multi-physics phenomena in metal additive manufacturing (AM) processes, yet accurate and robust predictions remain challenging. This book chapter summarizes the recent method development at UIUC for simulating thermal multiphase flows in laser powder bed fusion (LPBF) and directed energy deposition (DED) processes. Two main method developments are discussed. The first is a mixed interface-capturing/interface-tracking computational framework aiming to explicitly treat the gas-metal interface without mesh motion/re-meshing. The second is a physics-based and non-empirical deposit geometry model for DED processes. The proposed framework's accuracy is assessed by thoroughly comparing the simulated results against experimental measurements on various quantities. We also report critical quantities that experiments can not measure to show the predictive capability of the developed methods. 10.06.2022 16:32 Magnetizing laser-driven inertial fusion implosions Nuclear fusion is a widely studied process through which atomic nuclei of a low atomic number fuse together to form a heavier nucleus, while releasing a large amount of energy. Nuclear fusion reactions can be produced using a method known as inertial confinement fusion, which entails the use of powerful lasers to implode a fuel capsule and produce plasma. 05:13 Transport in fusion plasmas: is the tail wagging the dog?. (arXiv:2206.04193v1 [physics.plasm-ph]) Turbulent plasmas notably self-organize to higher energy states upon application of additional free energy sources or modification of edge operating conditions. Mechanisms whereby such bifurcations occur have been actively debated for decades. Enhanced confinement occurs at the plasma edge, where a shortfall of predicted turbulence intensity has been puzzling scientists for decades. We show, from the primitive kinetic equations that both problems are connected and that interplay of confined plasma turbulence with its material boundaries is essential to curing the shortfall of predicted turbulence and to triggering spontaneous transport barrier onset at the plasma edge. Both problems determine access to improved confinement and are central to fusion research. A comprehensive discussion of the underlying mechanisms is proposed. These results, highly relevant to the quest for magnetic fusion may also be generic to many problems in fluids and plasmas where turbulence self-advection is 05:12 Classification of COVID-19 in Chest X-ray Images Using Fusion of Deep Features and LightGBM. (arXiv:2206.04548v1 [eess.IV]) The COVID-19 disease was first discovered in Wuhan, China, and spread quickly worldwide. After the COVID-19 pandemic, many researchers have begun to identify a way to diagnose the COVID-19 using chest X-ray images. The early diagnosis of this disease can significantly impact the treatment process. In this article, we propose a new technique that is faster and more accurate than the other methods reported in the literature. The proposed method uses a combination of DenseNet169 and MobileNet Deep Neural Networks to extract the features of the patient's X-ray images. Using the univariate feature selection algorithm, we refined the features for the most important ones. Then we applied the selected features as input to the LightGBM (Light Gradient Boosting Machine) algorithm for classification. To assess the effectiveness of the proposed method, the ChestX-ray8 dataset, which includes 1125 X-ray images of the patient's chest, was used. The proposed method achieved 98.54% and 91.11% 09.06.2022 19:33 New feedback system can improve efficiency of fusion reactions Scientists at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) have refined the use of magnetic fields to improve the performance of doughnut-shaped fusion facilities known as tokamaks. The improved technique protects internal parts from damage by instabilities called "edge-localized modes" (ELMs) and allows tokamaks to operate for longer without pausing. 08.06.2022 20:12 Uncovering a novel way to bring the energy that powers the sun and stars to Earth Scientists at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) have uncovered critical new details about fusion facilities that use lasers to compress the fuel that produces fusion energy. The new data could help lead to the improved design of future laser facilities that harness the fusion process that drives the sun and stars. 10:42 Collaborative Intelligence Orchestration: Inconsistency-Based Fusion of Semi-Supervised Learning and Active Learning. (arXiv:2206.03288v1 [cs.LG]) While annotating decent amounts of data to satisfy sophisticated learning models can be cost-prohibitive for many real-world applications. Active learning (AL) and semi-supervised learning (SSL) are two effective, but often isolated, means to alleviate the data-hungry problem. Some recent studies explored the potential of combining AL and SSL to better probe the unlabeled data. However, almost all these contemporary SSL-AL works use a simple combination strategy, ignoring SSL and AL's inherent relation. Further, other methods suffer from high computational costs when dealing with large-scale, high-dimensional datasets. Motivated by the industry practice of labeling data, we propose an innovative Inconsistency-based virtual aDvErsarial Active Learning (IDEAL) algorithm to further investigate SSL-AL's potential superiority and achieve mutual enhancement of AL and SSL, i.e., SSL propagates label information to unlabeled samples and provides smoothed embeddings for AL, while AL excludes 07.06.2022 06:13 Machine Learning for Detection of 3D Features using sparse X-ray data. (arXiv:2206.02564v1 [cs.CV]) In many inertial confinement fusion experiments, the neutron yield and other parameters cannot be completely accounted for with one and two dimensional models. This discrepancy suggests that there are three dimensional effects which may be significant. Sources of these effects include defects in the shells and shell interfaces, the fill tube of the capsule, and the joint feature in double shell targets. Due to their ability to penetrate materials, X-rays are used to capture the internal structure of objects. Methods such as Computational Tomography use X-ray radiographs from hundreds of projections in order to reconstruct a three dimensional model of the object. In experimental environments, such as the National Ignition Facility and Omega-60, the availability of these views is scarce and in many cases only consist of a single line of sight. Mathematical reconstruction of a 3D object from sparse views is an ill-posed inverse problem. These types of problems are typically solved by 06:13 An Assessment Of Full Wave Effects On Maxwellian Lower Hybrid Wave Damping. (arXiv:2206.01773v1 [physics.plasm-ph]) Lower-hybrid current drive (LHCD) actuators are important components of modern day fusion experiments as well as proposed fusion reactors. However, simulations of LHCD often differ substantially from experimental results, and from each other, especially in the inferred power deposition profile shape. Here we investigate some possible causes of this discrepancy; "full-wave" effects such as interference and diffraction, which are omitted from standard raytracing simulations and the breakdown of the raytracing near reflections and caustics. We compare raytracing simulations to state-of-the-art full-wave simulations using matched hot-plasma dielectric tensors in realistic tokamak scenarios for the first time. We show that differences between full-wave simulations and raytracing in previous work were primarily due to numerical and physical inconsistencies in the simulations, and we demonstrate that good agreement between raytracing and full-wave simulations can be obtained in reactor 06:13 M2FNet: Multi-modal Fusion Network for Emotion Recognition in Conversation. (arXiv:2206.02187v1 [cs.CV]) Emotion Recognition in Conversations (ERC) is crucial in developing sympathetic human-machine interaction. In conversational videos, emotion can be present in multiple modalities, i.e., audio, video, and transcript. However, due to the inherent characteristics of these modalities, multi-modal ERC has always been considered a challenging undertaking. Existing ERC research focuses mainly on using text information in a discussion, ignoring the other two modalities. We anticipate that emotion recognition accuracy can be improved by employing a multi-modal approach. Thus, in this study, we propose a Multi-modal Fusion Network (M2FNet) that extracts emotion-relevant features from visual, audio, and text modality. It employs a multi-head attention-based fusion mechanism to combine emotion-rich latent representations of the input data. We introduce a new feature extractor to extract latent features from the audio and visual modality. The proposed feature extractor is trained with a novel 06.06.2022 22:43 Experts chip away at corrosion for the future of fusion Practical fusion energy is not just a dream at the Department of Energy’s Oak Ridge National Laboratory. Experts 16:33 Experts chip away at corrosion for the future of fusion Practical fusion energy is not just a dream at the Department of Energy's Oak Ridge National Laboratory. Experts in fusion and material science are working together to develop solutions that will make a fusion pilot plant—and ultimately carbon-free, abundant fusion electricity—possible. 14:42 Experts chip away at corrosion for the future of fusion Practical fusion energy is not just a dream at the Department of Energy’s Oak Ridge National Laboratory. Experts 09:12 Unsupervised Discovery of Non-Linear Plasma Physics using Differentiable Kinetic Simulations. (arXiv:2206.01637v1 [physics.plasm-ph]) Plasma supports collective modes and particle-wave interactions that leads to complex behavior in, for example, inertial fusion energy applications. While plasma can sometimes be modeled as a charged fluid, a kinetic description is often crucial for studying nonlinear effects in the higher dimensional momentum-position phase-space that describes the full complexity of plasma dynamics. We create a differentiable solver for the 3D partial-differential-equation describing the plasma kinetics and introduce a domain-specific objective function. Using this framework, we perform gradient-based optimization of neural networks that provide forcing function parameters to the differentiable solver given a set of initial conditions. We apply this to an inertial-fusion relevant configuration and find that the optimization process exploits a novel physical effect. 03.06.2022 10:03 Runaway dynamics in disruptions with current relaxation. (arXiv:2206.00904v1 [physics.plasm-ph]) The safe operation of tokamak reactors requires a reliable modeling capability of disruptions, and in particular the spatio-temporal dynamics of associated runaway electron currents. In a disruption, instabilities can break up magnetic surfaces into chaotic field line regions, causing current profile relaxation, as well as a rapid radial transport of heat and particles. Using a mean-field helicity transport model implemented in the disruption runaway modeling framework DREAM, we calculate the dynamics of runaway electrons in the presence of current relaxation events. In scenarios where flux surfaces remain intact in parts of the plasma, a skin current is induced at the boundary of the intact magnetic field region. This skin current region becomes an important center concerning the subsequent dynamics: It may turn into a hot ohmic current channel, or a sizable radially localized runaway beam, depending on the heat transport. If the intact region is in the plasma edge, runaway generation 10:03 Musical Instrument Recognition by XGBoost Combining Feature Fusion. (arXiv:2206.00901v1 [cs.SD]) Musical instrument classification is one of the focuses of Music Information Retrieval (MIR). In order to solve the problem of poor performance of current musical instrument classification models, we propose a musical instrument classification algorithm based on multi-channel feature fusion and XGBoost. Based on audio feature extraction and fusion of the dataset, the features are input into the XGBoost model for training; secondly, we verified the superior performance of the algorithm in the musical instrument classification task by com-paring different feature combinations and several classical machine learning models such as Naive Bayes. The algorithm achieves an accuracy of 97.65% on the Medley-solos-DB dataset, outperforming existing models. The experiments provide a reference for feature selection in feature engineering for musical instrument classification. 01.06.2022 18:53 Unprecedented level of insight into plasma edge phenomena Producing energy and heat using plasma fusion is one of the promising technologies for the transition to sustainable energy sources. One of the challenges is managing the temperatures in the plasma edge. Ph.D. researcher Artur Perek has built an imaging system known as MANTIS to image and monitor temperature in the plasma edge, and he has improved the software performance to enhance control of plasma edge temperatures. Perek defended his thesis at the department of Applied Physics on April 13th. 07:52 2022 Review of Data-Driven Plasma Science. (arXiv:2205.15832v1 [physics.plasm-ph]) Data science and technology offer transformative tools and methods to science. This review article highlights latest development and progress in the interdisciplinary field of data-driven plasma science (DDPS). A large amount of data and machine learning algorithms go hand in hand. Most plasma data, whether experimental, observational or computational, are generated or collected by machines today. It is now becoming impractical for humans to analyze all the data manually. Therefore, it is imperative to train machines to analyze and interpret (eventually) such data as intelligently as humans but far more efficiently in quantity. Despite the recent impressive progress in applications of data science to plasma science and technology, the emerging field of DDPS is still in its infancy. Fueled by some of the most challenging problems such as fusion energy, plasma processing of materials, and fundamental understanding of the universe through observable plasma phenomena, it is expected that 07:52 Thermodynamic Flux-Force Closure Relations for Systems out of the Onsager Region. (arXiv:2205.15315v1 [physics.plasm-ph]) The objective of this work is to determine the nonlinear flux-force relations for systems out of Onsager's region that respect the existing thermodynamic theorems for systems far from equilibrium. To this aim, a thermodynamic theory for irreversible processes [referred to as the Thermodynamical Field Theory (TFT)] has been developed. The TFT rests upon the concept of equivalence between thermodynamic systems: "The equivalent character of two alternative descriptions of a thermodynamic system is ensured if, and only if, the entropy production and the Glansdorff-Prigogine dissipative quantity remain unaltered under the thermodynamic forces transformation". The TCT leads naturally to the "Thermodynamic Covariance Principle" (TCP) stating that "The nonlinear closure equations, i.e., the flux-force relations, must be covariant under TCT". In this work, we provide the explicit expression of the nonlinear PDEs, subjected to the appropriate boundary conditions, which have to be satisfied by 00:30 Nuclear fusion is the holy grail of clean energy. We're closer to it than ever • Video: This is what determines the price of gas 31.05.2022 10:23 SuperVoice: Text-Independent Speaker Verification Using Ultrasound Energy in Human Speech. (arXiv:2205.14496v1 [cs.SD]) Voice-activated systems are integrated into a variety of desktop, mobile, and Internet-of-Things (IoT) devices. However, voice spoofing attacks, such as impersonation and replay attacks, in which malicious attackers synthesize the voice of a victim or simply replay it, have brought growing security concerns. Existing speaker verification techniques distinguish individual speakers via the spectrographic features extracted from an audible frequency range of voice commands. However, they often have high error rates and/or long delays. In this paper, we explore a new direction of human voice research by scrutinizing the unique characteristics of human speech at the ultrasound frequency band. Our research indicates that the high-frequency ultrasound components (e.g. speech fricatives) from 20 to 48 kHz can significantly enhance the security and accuracy of speaker verification. We propose a speaker verification system, SUPERVOICE that uses a two-stream DNN architecture with a feature fusion 30.05.2022 05:53 Resonant Ion Confinement Fusion Concept. (arXiv:2205.14013v1 [physics.gen-ph]) Based on the theorized possibilities of resonant ion confinement, for a Deuteron cloud in a Penning-Malmberg trap with a specially configured rotating wall, the opportunity to design a new type of fusion device is prospected. It is proven that, for some trap configurations, nuclear fusion reactions should take place and, in that case, Lawson's criterion for an efficient fusion reactor is met. Furthermore, the reactor could have a compact design and, since it should not require a large facility, it can function as a fusion cell with a pure ion thermal gas. 29.05.2022 16:21 Robotic snakes could keep Britain’s nuclear fusion hopes alive 28.05.2022 16:22 This government lab in Idaho is researching fusion, the 'holy grail' of clean energy, as billions pour into the space Fusion is getting billions in private investment, interest from young scientists. Here's a look at how the government is researching fuel cycles and safety. 16:12 This government lab in Idaho is researching fusion, the 'holy grail' of clean energy, as billions pour into the space Fusion is getting billions in private investment, interest from young scientists. Here's a look at how the government is researching fuel cycles and safety. 27.05.2022 06:43 Enhanced collisionless laser absorption in strongly magnetized plasmas. (arXiv:2205.13478v1 [physics.plasm-ph]) Strongly magnetizing a plasma adds a range of waves that do not exist in unmagnetized plasmas and enlarges the laser-plasma interaction (LPI) landscape. In this paper, we use particle-in-cell (PIC) simulations to investigate strongly magnetized LPI in one dimension under conditions relevant for magneto-inertial fusion experiments, focusing on a regime where the electron-cyclotron frequency is greater than the plasma frequency and the magnetic field is at an oblique angle with respect to the wave vectors. We show that when the electron cyclotron frequency is about half the laser frequency, the laser light resonantly decays to magnetized plasma waves via primary and secondary instabilities with large growth rates. These distinct magnetic-field-controlled instabilities, which we collectively call two-magnon decays, are analogous to two-plasmon decays in unmagnetized plasmas. Since the oblique magnetic field introduces additional phase mixing mechanisms during wave-particle interactions, 06:43 Cost-Optimal System Performance Maps for Laser-Accelerated Sailcraft. (arXiv:2205.13138v1 [astro-ph.IM]) Breakthrough Starshot is an initiative to explore the Centauri system using laser-accelerated sailcraft. Earlier work produced a point design for a 0.2 c mission carrying 1 g of payload. The present work widens the design space to missions having 0.1 mg to 100 kt payload and 0.0001-0.99 c (6-60,000 au/yr) cruise velocity. Also, the beam director may now draw up to 5 GW of power directly from the grid to augment the power drawn from its energy storage system. Augmenting stored energy with grid power shrinks beam director capital cost by 1-5 orders of magnitude. The wider design space encompasses new possibilities: A 0.1 mg microbiome accelerated to 0.01 c in only 2 min by a beam director that expends \$6k worth of energy. A 10 kg Solar system cubesat accelerated to 0.001 c (60 au/yr) by a \$600M beam director that expends \$60M worth of energy per mission. A progression from cost-optimized point designs to whole performance maps has been made possible by replacing numerical trajectory

06:42 Flying Hydraulically Amplified Electrostatic Gripper System for Aerial Object Manipulation. (arXiv:2205.13011v1 [cs.RO])

Rapid and versatile object manipulation in air is an open challenge. An energy-efficient and adaptive soft gripper combined with an agile aerial vehicle could revolutionize aerial robotic manipulation in areas such as warehousing. This paper presents a bio-inspired gripper powered by hydraulically amplified electrostatic actuators mounted to a quadcopter that can interact safely and naturally with its environment. Our gripping concept is motivated by an eagle's talon. Our custom multi-actuator type is inspired by a previous scorpion tail design (consisting of a base electrode and pouches stacked adjacently) and spider-inspired joints (classic pouch motors with a flexible hinge layer). A fusion of these two concepts realizes a higher force output than single-actuator types under considerable deflections of up to 25{\deg}. By adding a sandwich hinge layer structure to the classic pouch motor concept we improve the overall robustness of the gripper. We show, for the first time, that soft

26.05.2022
09:32 Investigation of atomic and molecular rates in plasma-edge simulations through experiment-simulation comparisons. (arXiv:2205.12715v1 [physics.plasm-ph])

Divertor plasma detachment is likely needed for the function of magnetically confined nuclear fusion. It greatly reduces the particle and heat flux incident on a target, and thus reduces the sputtering and heat loading on the target. It is therefore advantageous to have accurate simulations of plasma behaviour in the divertor to design future Tokamak divertors. A common simulation package used for this purpose is SOLPS-ITER, which has previously been observed to be underestimating the contribution of molecular effects to plasma detachment. To correct this, its reaction rate for molecular charge exchange was altered. This alteration resulted in a significant increase in the density of $D_2^+$ ions within the divertor. The particle balance and $D\alpha$ emission data was then computed from the simulation outputs. These were compared to experimental data and a set of post-processing routines for the original simulations that predicts the effect of a corrected charge exchange rate. In

25.05.2022
18:33 Physicists just rewrote a foundational rule for nuclear fusion reactors that could unleash twice the power

Future fusion reactions inside tokamaks could shine even brighter than before, thanks to groundbreaking new research to find the maximum density of the hydrogen plasma fuel that powers them.

17:23 Fusion experts tackle cooling strategies for fusion fuel cycle

To achieve practical energy from fusion, extreme heat from the fusion system "blanket" component must be extracted safely and efficiently. Oak Ridge National Laboratory fusion experts are exploring how tiny 3D-printed obstacles placed inside the narrow pipes of a custom-made cooling system could be a solution for removing heat from the blanket.

05:06 SparseLNR: Accelerating Sparse Tensor Computations Using Loop Nest Restructuring. (arXiv:2205.11622v1 [cs.PL])

Sparse tensor algebra computations have become important in many real-world applications like machine learning, scientific simulations, and data mining. Hence, automated code generation and performance optimizations for tensor algebra kernels are paramount. Recent advancements such as the Tensor Algebra Compiler (TACO) greatly generalize and automate the code generation for tensor algebra expressions. However, the code generated by TACO for many important tensor computations remains suboptimal due to the absence of a scheduling directive to support transformations such as distribution/fusion. This paper extends TACO's scheduling space to support kernel distribution/loop fusion in order to reduce asymptotic time complexity and improve locality of complex tensor algebra computations. We develop an intermediate representation (IR) for tensor operations called branched iteration graph which specifies breakdown of the computation into smaller ones (kernel distribution) and then fuse

24.05.2022
08:02 Exact surface energy of the $D^{(1)}_2$ spin chain with generic non-diagonal boundary reflections. (arXiv:2205.10818v1 [math-ph])

The exact solution of the $D^{(1)}_2$ quantum spin chain with generic non-diagonal boundary reflections is obtained. It is found that the generating functional of conserved quantities of the system can be factorized as the product of transfer matrices of two anisotropic $XXZ$ spin chains with open boundary conditions. By using the factorization identities and the fusion technique, the eigenvalues and the Bethe ansatz equations of the model are obtained. The eigenvalues are also parameterized by the zero roots of the transfer matrix, and the patterns of root distributions are obtained. Based on them, ground states energy and the surface energies induced by the twisted boundary magnetic fields in the thermodynamic limit are obtained. These results are checked by the numerical calculations. The corresponding isotropic limit is also discussed. The results given in this paper are the foundation to study the exact physical properties of high rank $D^{(1)}_{n}$ model by using the nested

08:02 Deep Feature Fusion via Graph Convolutional Network for Intracranial Artery Labeling. (arXiv:2205.10757v1 [eess.IV])

Intracranial arteries are critical blood vessels that supply the brain with oxygenated blood. Intracranial artery labels provide valuable guidance and navigation to numerous clinical applications and disease diagnoses. Various machine learning algorithms have been carried out for automation in the anatomical labeling of cerebral arteries. However, the task remains challenging because of the high complexity and variations of intracranial arteries. This study investigates a novel graph convolutional neural network with deep feature fusion for cerebral artery labeling. We introduce stacked graph convolutions in an encoder-core-decoder architecture, extracting high-level representations from graph nodes and their neighbors. Furthermore, we efficiently aggregate intermediate features from different hierarchies to enhance the proposed model's representation capability and labeling performance. We perform extensive experiments on public datasets, in which the results prove the superiority of

23.05.2022
10:43 Deep electric field predictions by drift-reduced Braginskii theory with plasma-neutral interactions based upon experimental images of boundary turbulence. (arXiv:2204.11689v1 [physics.plasm-ph] CROSS LISTED)

We present 2-dimensional turbulent electric field calculations via physics-informed deep learning consistent with (i) drift-reduced Braginskii theory under the framework of an axisymmetric fusion plasma with purely toroidal field and (ii) experimental estimates of the fluctuating electron density and temperature obtained from analysis of gas puff imaging of a discharge on the Alcator C-Mod tokamak. The inclusion of effects from the locally puffed atomic helium on particle and energy sources within the reduced plasma turbulence model are found to strengthen correlations between the electric field and electron pressure. The neutrals are also directly associated with an observed broadening in the distribution of turbulent field amplitudes and increased ${\bf E \times B}$ shearing rates.

10:43 Deep electric field predictions by drift-reduced Braginskii theory with plasma-neutral interactions based upon experimental images of boundary turbulence. (arXiv:2204.11689v1 [physics.plasm-ph] CROSS LISTED)

We present 2-dimensional turbulent electric field calculations via physics-informed deep learning consistent with (i) drift-reduced Braginskii theory under the framework of an axisymmetric fusion plasma with purely toroidal field and (ii) experimental estimates of the fluctuating electron density and temperature obtained from analysis of gas puff imaging of a discharge on the Alcator C-Mod tokamak. The inclusion of effects from the locally puffed atomic helium on particle and energy sources within the reduced plasma turbulence model are found to strengthen correlations between the electric field and electron pressure. The neutrals are also directly associated with an observed broadening in the distribution of turbulent field amplitudes and increased ${\bf E \times B}$ shearing rates.

10:43 Constructive Interpretability with CoLabel: Corroborative Integration, Complementary Features, and Collaborative Learning. (arXiv:2205.10011v1 [cs.CV])

Machine learning models with explainable predictions are increasingly sought after, especially for real-world, mission-critical applications that require bias detection and risk mitigation. Inherent interpretability, where a model is designed from the ground-up for interpretability, provides intuitive insights and transparent explanations on model prediction and performance. In this paper, we present CoLabel, an approach to build interpretable models with explanations rooted in the ground truth. We demonstrate CoLabel in a vehicle feature extraction application in the context of vehicle make-model recognition (VMMR). CoLabel performs VMMR with a composite of interpretable features such as vehicle color, type, and make, all based on interpretable annotations of the ground truth labels. First, CoLabel performs corroborative integration to join multiple datasets that each have a subset of desired annotations of color, type, and make. Then, CoLabel uses decomposable branches to extract

10:43 Subcellular Protein Localisation in the Human Protein Atlas using Ensembles of Diverse Deep Architectures. (arXiv:2205.09841v1 [cs.CV])

Automated visual localisation of subcellular proteins can accelerate our understanding of cell function in health and disease. Despite recent advances in machine learning (ML), humans still attain superior accuracy by using diverse clues. We show how this gap can be narrowed by addressing three key aspects: (i) automated improvement of cell annotation quality, (ii) new Convolutional Neural Network (CNN) architectures supporting unbalanced and noisy data, and (iii) informed selection and fusion of multiple & diverse machine learning models. We introduce a new "AI-trains-AI" method for improving the quality of weak labels and propose novel CNN architectures exploiting wavelet filters and Weibull activations. We also explore key factors in the multi-CNN ensembling process by analysing correlations between image-level and cell-level predictions. Finally, in the context of the Human Protein Atlas, we demonstrate that our system achieves state-of-the-art performance in the multi-label

20.05.2022
04:13 Turbulent field fluctuations in gyrokinetic and fluid plasmas. (arXiv:2107.09744v2 [physics.plasm-ph] CROSS LISTED)

A key uncertainty in the design and development of magnetic confinement fusion energy reactors is predicting edge plasma turbulence. An essential step in overcoming this uncertainty is the validation in accuracy of reduced turbulent transport models. Drift-reduced Braginskii two-fluid theory is one such set of reduced equations that has for decades simulated boundary plasmas in experiment, but significant questions exist regarding its predictive ability. To this end, using a novel physics-informed deep learning framework, we demonstrate the first ever direct quantitative comparisons of turbulent field fluctuations between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling with good overall agreement found in magnetized helical plasmas at low normalized pressure. This framework is readily adaptable to experimental and astrophysical environments, and presents a new technique for the numerical validation and discovery of reduced global plasma turbulence models.

04:13 Atomistic simulations of nanoindentation in single crystalline tungsten: The role of interatomic potentials. (arXiv:2205.09165v1 [physics.comp-ph])

The design of the next generation of nuclear fusion machines needs efficient Plasma Facing materials (PFMs) that can withstand extreme operating conditions due to the direct interaction with the fusion plasma, and BCC metals can fulfill these requirements, in particular tungsten. However, the understanding of the behavior of these materials at extreme operating conditions, such as irradiation and high temperatures, also depends on the capacity for efficient molecular simulations using appropriate interatomic potentials. In this work, we perform Molecular dynamics (MD) simulations to emulate experimental nanoindentation tests of crystalline pure [111] W by two different Embedded Atom Method(EAM)-based interatomic potentials for describing the interaction of W-W and W-H/W-W, respectively. The characterization of W mechanical properties is done by a detailed analysis of the dislocation nucleation and evolution during the early stages of the elastic to plastic deformation transition.

04:13 Turbulent field fluctuations in gyrokinetic and fluid plasmas. (arXiv:2107.09744v2 [physics.plasm-ph] CROSS LISTED)

A key uncertainty in the design and development of magnetic confinement fusion energy reactors is predicting edge plasma turbulence. An essential step in overcoming this uncertainty is the validation in accuracy of reduced turbulent transport models. Drift-reduced Braginskii two-fluid theory is one such set of reduced equations that has for decades simulated boundary plasmas in experiment, but significant questions exist regarding its predictive ability. To this end, using a novel physics-informed deep learning framework, we demonstrate the first ever direct quantitative comparisons of turbulent field fluctuations between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling with good overall agreement found in magnetized helical plasmas at low normalized pressure. This framework is readily adaptable to experimental and astrophysical environments, and presents a new technique for the numerical validation and discovery of reduced global plasma turbulence models.

04:13 CORPS: Cost-free Rigorous Pseudo-labeling based on Similarity-ranking for Brain MRI Segmentation. (arXiv:2205.09601v1 [cs.CV])

Segmentation of brain magnetic resonance images (MRI) is crucial for the analysis of the human brain and diagnosis of various brain disorders. The drawbacks of time-consuming and error-prone manual delineation procedures are aimed to be alleviated by atlas-based and supervised machine learning methods where the former methods are computationally intense and the latter methods lack a sufficiently large number of labeled data. With this motivation, we propose CORPS, a semi-supervised segmentation framework built upon a novel atlas-based pseudo-labeling method and a 3D deep convolutional neural network (DCNN) for 3D brain MRI segmentation. In this work, we propose to generate expert-level pseudo-labels for unlabeled set of images in an order based on a local intensity-based similarity score to existing labeled set of images and using a novel atlas-based label fusion method. Then, we propose to train a 3D DCNN on the combination of expert and pseudo labeled images for binary segmentation

04:13 Twist Decoding: Diverse Generators Guide Each Other. (arXiv:2205.09273v1 [cs.CL])

Natural language generation technology has recently seen remarkable progress with large-scale training, and many natural language applications are now built upon a wide range of generation models. Combining diverse models may lead to further progress, but conventional ensembling (e.g., shallow fusion) requires that they share vocabulary/tokenization schemes. We introduce Twist decoding, a simple and general inference algorithm that generates text while benefiting from diverse models. Our method does not assume the vocabulary, tokenization or even generation order is shared. Our extensive evaluations on machine translation and scientific paper summarization demonstrate that Twist decoding substantially outperforms each model decoded in isolation over various scenarios, including cases where domain-specific and general-purpose models are both available. Twist decoding also consistently outperforms the popular reranking heuristic where output candidates from one model is rescored by

18.05.2022
09:25 DouFu: A Double Fusion Joint Learning Method For Driving Trajectory Representation. (arXiv:2205.08356v1 [cs.LG])

Driving trajectory representation learning is of great significance for various location-based services, such as driving pattern mining and route recommendation. However, previous representation generation approaches tend to rarely address three challenges: 1) how to represent the intricate semantic intentions of mobility inexpensively; 2) complex and weak spatial-temporal dependencies due to the sparsity and heterogeneity of the trajectory data; 3) route selection preferences and their correlation to driving behavior. In this paper, we propose a novel multimodal fusion model, DouFu, for trajectory representation joint learning, which applies multimodal learning and attention fusion module to capture the internal characteristics of trajectories. We first design movement, route, and global features generated from the trajectory data and urban functional zones and then analyze them respectively with the attention encoder or feed forward network. The attention fusion module incorporates

17.05.2022
20:15 A new law unchains fusion energy

Physicists at EPFL, within a large European collaboration, have revised one of the fundamental laws that has been foundational to plasma and fusion research for over three decades, even governing the design of megaprojects like ITER. The update shows that we can actually safely use more hydrogen fuel in fusion reactors, and therefore obtain more energy than previously thought.

05:13 Physics-informed machine learning techniques for edge plasma turbulence modelling in computational theory and experiment. (arXiv:2205.07838v1 [physics.plasm-ph])

Edge plasma turbulence is critical to the performance of magnetic confinement fusion devices. Towards better understanding edge turbulence in both theory and experiment, a custom-built physics-informed deep learning framework constrained by partial differential equations is developed to accurately learn turbulent fields consistent with the two-fluid theory from partial observations of electron pressure. This calculation is not otherwise possible using conventional equilibrium models. With this technique, the first direct quantitative comparisons of turbulent fields between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling are demonstrated with good overall agreement found in magnetized helical plasmas at low normalized pressure. To translate these computational techniques to experimental fusion plasmas, a novel method to translate brightness measurements of HeI line radiation into local plasma fluctuations is demonstrated via a newly created deep learning

05:12 Physics-informed machine learning techniques for edge plasma turbulence modelling in computational theory and experiment. (arXiv:2205.07838v1 [physics.plasm-ph])

Edge plasma turbulence is critical to the performance of magnetic confinement fusion devices. Towards better understanding edge turbulence in both theory and experiment, a custom-built physics-informed deep learning framework constrained by partial differential equations is developed to accurately learn turbulent fields consistent with the two-fluid theory from partial observations of electron pressure. This calculation is not otherwise possible using conventional equilibrium models. With this technique, the first direct quantitative comparisons of turbulent fields between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling are demonstrated with good overall agreement found in magnetized helical plasmas at low normalized pressure. To translate these computational techniques to experimental fusion plasmas, a novel method to translate brightness measurements of HeI line radiation into local plasma fluctuations is demonstrated via a newly created deep learning

05:12 Physics-informed machine learning techniques for edge plasma turbulence modelling in computational theory and experiment. (arXiv:2205.07838v1 [physics.plasm-ph])

Edge plasma turbulence is critical to the performance of magnetic confinement fusion devices. Towards better understanding edge turbulence in both theory and experiment, a custom-built physics-informed deep learning framework constrained by partial differential equations is developed to accurately learn turbulent fields consistent with the two-fluid theory from partial observations of electron pressure. This calculation is not otherwise possible using conventional equilibrium models. With this technique, the first direct quantitative comparisons of turbulent fields between electrostatic two-fluid theory and electromagnetic gyrokinetic modelling are demonstrated with good overall agreement found in magnetized helical plasmas at low normalized pressure. To translate these computational techniques to experimental fusion plasmas, a novel method to translate brightness measurements of HeI line radiation into local plasma fluctuations is demonstrated via a newly created deep learning

01:52 Bernard Bigot, head of gigantic ITER fusion project, dies

Observers credit him with reforming the world’s largest science project

15.05.2022
13:33 French scientist leading nuclear fusion project dies at 72

Bernard Bigot, a French scientist leading a vast international effort to demonstrate that nuclear fusion can be a viable source of energy, has died. He was 72.

13.05.2022
08:43 A single-field-period quasi-isodynamic stellarator. (arXiv:2205.05797v1 [physics.plasm-ph])

A single-field-period quasi-isodynamic stellarator configuration is presented. This configuration, which resembles a twisted strip, is obtained by the method of direct construction, that is, it is found via an expansion in the distance from the magnetic axis. Its discovery, however, relied on an additional step involving numerical optimization, performed within the space of near-axis configurations defined by a set of adjustable magnetic-field parameters. This optimization, completed in 30 seconds on a single cpu core using the SIMSOPT code, yields a solution with excellent confinement, as measured by the conventional figure of merit for neoclassical transport, effective ripple, at a modest aspect ratio of eight. The optimization parameters that led to this configuration are described, its confinement properties are assessed, and a set of magnetic-field coils is found. The resulting transport at low collisionality is much smaller than that of W7-X, and the device needs significantly

08:43 Processing of massive Rutherford Back-scattering Spectrometry data by artificial neural networks. (arXiv:2205.05774v1 [physics.comp-ph])

Rutherford Backscattering Spectrometry (RBS) is an important technique providing elemental information of the near surface region of samples with high accuracy and robustness. However, this technique lacks throughput by the limited rate of data processing and is hardly routinely applied in research with a massive number of samples (i.e. hundreds or even thousands of samples). The situation is even worse for complex samples. If roughness or porosity is present in those samples the simulation of such structures is computationally demanding. Fortunately, Artificial Neural Networks (ANN) show to be a great ally for massive data processing of ion beam data. In this paper, we report the performance comparison of ANN against human evaluation and an automatic fit routine running on batch mode. 500 spectra of marker layers from the stellarator W7-X were used as study case. The results showed ANN as more accurate than humans and more efficient than automatic fits.

12.05.2022
12:02 Ukraine's Kalush Orchestra throws down motherland's beats at Eurovision

Tapping traditional Ukrainian folk music but mashing up an invigorating hiphop beat with a haunting, lullaby refrain, "Stefania" was written last year by the band's frontman, 27-year-old rapper Oleh Psiuk, as a tribute to his mother. But the song selected to represent Ukraine at Eurovision -- just days before Russia's invasion -- has taken on outsized meaning for a country nearing its third month of war. It contains nostalgic lyrics such as "I'll always find my way home even if all the roads are destroyed" and celebrates cultural identity and the motherland. Standing out in the competition long cheered for its flamboyance and camp, the band received a standing ovation on Tuesday after passing the semifinals. It is considered by bookmakers a favourite to become Eurovision's outright winner at the finale on Saturday. "My mum is in Ukraine and many of my relatives are in Ukraine but there is really no safe place in Ukraine at the moment," Psiuk told AFP through an interpreter. "It's

04:32 CNN-LSTM Based Multimodal MRI and Clinical Data Fusion for Predicting Functional Outcome in Stroke Patients. (arXiv:2205.05545v1 [eess.IV])

Clinical outcome prediction plays an important role in stroke patient management. From a machine learning point-of-view, one of the main challenges is dealing with heterogeneous data at patient admission, i.e. the image data which are multidimensional and the clinical data which are scalars. In this paper, a multimodal convolutional neural network - long short-term memory (CNN-LSTM) based ensemble model is proposed. For each MR image module, a dedicated network provides preliminary prediction of the clinical outcome using the modified Rankin scale (mRS). The final mRS score is obtained by merging the preliminary probabilities of each module dedicated to a specific type of MR image weighted by the clinical metadata, here age or the National Institutes of Health Stroke Scale (NIHSS). The experimental results demonstrate that the proposed model surpasses the baselines and offers an original way to automatically encode the spatio-temporal context of MR images in a deep learning

10.05.2022
09:43 Measuring Cognitive Workload Using Multimodal Sensors. (arXiv:2205.04235v1 [q-bio.NC])

This study aims to identify a set of indicators to estimate cognitive workload using a multimodal sensing approach and machine learning. A set of three cognitive tests were conducted to induce cognitive workload in twelve participants at two levels of task difficulty (Easy and Hard). Four sensors were used to measure the participants' physiological change, including, Electrocardiogram (ECG), electrodermal activity (EDA), respiration (RESP), and blood oxygen saturation (SpO2). To understand the perceived cognitive workload, NASA-TLX was used after each test and analysed using Chi-Square test. Three well-know classifiers (LDA, SVM, and DT) were trained and tested independently using the physiological data. The statistical analysis showed that participants' perceived cognitive workload was significantly different (p<0.001) between the tests, which demonstrated the validity of the experimental conditions to induce different cognitive levels. Classification results showed that a fusion of

09:43 Training and Upgrading Tokamak Power Plants with Remountable Superconducting Magnets. (arXiv:2205.04441v1 [physics.plasm-ph])

All high field superconductors producing magnetic fields above 12 T are brittle. Nevertheless, they will probably be the materials of choice in commercial tokamaks because the fusion power density in a tokamak scales as the fourth power of magnetic field. Here we propose using robust, ductile superconductors during the reactor commissioning phase in order to avoid brittle magnet failure while operational safety margins are being established. Here we use the PROCESS systems code to inform development strategy and to provide detailed capital-cost-minimised tokamak power plant designs. We propose building a 'demonstrator' tokamak with an electric power output of 100 MWe, a plasma fusion gain Qplasma = 17, a net gain Qnet = 1.3, a cost of electricity (COE) of \$1148 (2021 US) per MWh (at 75 % availability) and high temperature superconducting operational TF magnets producing 5.4 T on-axis and 12.5 T peak-field. It uses Nb-Ti training magnets and will cost about \$ 9.75 Bn (2021 US). An

09:43 Turbulent transport regimes in the tokamak boundary and operational limits. (arXiv:2205.04409v1 [physics.plasm-ph])

Two-fluid, three-dimensional, flux-driven, global, electromagnetic turbulence simulations carried out by using the GBS code are used to identify the main parameters controlling turbulent transport in the tokamak boundary and to delineate an electromagnetic phase space of edge turbulence. Four turbulent transport regimes are identified: (i) a regime of fully developed turbulence appearing at intermediate values of collisionality and $\beta$, with turbulence driven by resistive ballooning modes, related to the L-mode operation of tokamaks, (ii) a regime of reduced turbulent transport at low collisionality and large heat source, with turbulence driven by drift-waves, related to a high-density H-mode regime, (iii) a regime of extremely large turbulent transport at high collisionality, which is associated with the crossing of the density limit, and (iv) a regime above the ideal ballooning limit at high $\beta$, with global modes affecting the dynamics of the entire confined region, which

09:43 SELF-CARE: Selective Fusion with Context-Aware Low-Power Edge Computing for Stress Detection. (arXiv:2205.03974v1 [eess.SP])

Detecting human stress levels and emotional states with physiological body-worn sensors is a complex task, but one with many health-related benefits. Robustness to sensor measurement noise and energy efficiency of low-power devices remain key challenges in stress detection. We propose SELFCARE, a fully wrist-based method for stress detection that employs context-aware selective sensor fusion that dynamically adapts based on data from the sensors. Our method uses motion to determine the context of the system and learns to adjust the fused sensors accordingly, improving performance while maintaining energy efficiency. SELF-CARE obtains state-of-the-art performance across the publicly available WESAD dataset, achieving 86.34% and 94.12% accuracy for the 3-class and 2-class classification problems, respectively. Evaluation on real hardware shows that our approach achieves up to 2.2x (3-class) and 2.7x (2-class) energy efficiency compared to traditional sensor fusion.

09.05.2022
05:23 Optimization of quasisymmetric stellarators with self-consistent bootstrap current and energetic particle confinement. (arXiv:2205.02914v1 [physics.plasm-ph])

Quasisymmetry can greatly improve the confinement of energetic particles and thermal plasma in a stellarator. The magnetic field of a quasisymmetric stellarator at high plasma pressure is significantly affected by the bootstrap current, but the computational cost of accurate stellarator bootstrap calculations has precluded use inside optimization. Here, a new efficient method is demonstrated for optimization of quasisymmetric stellarator configurations such that the bootstrap current profile is consistent with the geometry. The approach is based on the fact that all neoclassical phenomena in quasisymmetry are isomorphic to those in axisymmetry. Therefore accurate formulae for the bootstrap current in tokamaks, which can be evaluated rapidly, can be applied also in stellarators. The deviation between this predicted parallel current and the actual parallel current in the magnetohydrodynamic equilibrium is penalized in the objective function, and the current profile of the equilibrium is

05:22 Quantifying Synthesis and Fusion and their Impact on Machine Translation. (arXiv:2205.03369v1 [cs.CL])

Theoretical work in morphological typology offers the possibility of measuring morphological diversity on a continuous scale. However, literature in Natural Language Processing (NLP) typically labels a whole language with a strict type of morphology, e.g. fusional or agglutinative. In this work, we propose to reduce the rigidity of such claims, by quantifying morphological typology at the word and segment level. We consider Payne (2017)'s approach to classify morphology using two indices: synthesis (e.g. analytic to polysynthetic) and fusion (agglutinative to fusional). For computing synthesis, we test unsupervised and supervised morphological segmentation methods for English, German and Turkish, whereas for fusion, we propose a semi-automatic method using Spanish as a case study. Then, we analyse the relationship between machine translation quality and the degree of synthesis and fusion at word (nouns and verbs for English-Turkish, and verbs in English-Spanish) and segment level

05:22 Fusion: Efficient and Secure Inference Resilient to Malicious Server and Curious Clients. (arXiv:2205.03040v1 [cs.CR])

In secure machine learning inference, most current schemes assume that the server is semi-honest and honestly follows the protocol but attempts to infer additional information. However, in real-world scenarios, the server may behave maliciously, e.g., using low-quality model parameters as inputs or deviating from the protocol. Although a few studies consider the security against the malicious server, they do not guarantee the model accuracy while preserving the privacy of both server's model and the client's inputs. Furthermore, a curious client may perform model extraction attacks to steal the server's model. To address these issues, we propose Fusion, an efficient and privacy-preserving inference scheme that is secure against the malicious server, and a curious client who may perform model extraction attacks. Without leveraging expensive cryptographic techniques, a novel mix-and-check method is designed to ensure that the server uses a well-trained model as input and correctly

06.05.2022
20:42 Implementation and reconfiguration of magnetic skyrmions-based logic gates in one single nanotrack

In one single nanotrack, a research team has achieved the annihilation, fusion and shunting of two skyrmions with opposite chirality via local reversal of the DMI, as well as the pinning effect of energy barriers on skyrmions.

06:13 On the Turbulent Behavior of a Magnetically Confined Plasma Near the X-Point. (arXiv:2205.02559v1 [physics.plasm-ph])

We construct a model for the turbulence near the X-point of a Tokamak device and, under suitable assumptions, we arrive to a closed equation for the electric field potential fluctuations. The analytical and numerical analysis is focused on a reduced two-dimensional formulation of the dynamics, which allows a direct mapping to the incompressible Navier-Stokes equation. The main merit of this study is to outline how the turbulence near the X-point, in correspondence to typical operation conditions of medium and large size Tokamaks, is dominated by the enstrophy cascade from large to smaller spatial scales.

06:13 Multi-Graph based Multi-Scenario Recommendation in Large-scale Online Video Services. (arXiv:2205.02446v1 [cs.AI])

Recently, industrial recommendation services have been boosted by the continual upgrade of deep learning methods. However, they still face de-biasing challenges such as exposure bias and cold-start problem, where circulations of machine learning training on human interaction history leads algorithms to repeatedly suggest exposed items while ignoring less-active ones. Additional problems exist in multi-scenario platforms, e.g. appropriate data fusion from subsidiary scenarios, which we observe could be alleviated through graph structured data integration via message passing. In this paper, we present a multi-graph structured multi-scenario recommendation solution, which encapsulates interaction data across scenarios with multi-graph and obtains representation via graph learning. Extensive offline and online experiments on real-world datasets are conducted where the proposed method demonstrates an increase of 0.63% and 0.71% in CTR and Video Views per capita on new users over

05.05.2022
14:31 How Toronto’s mainstay Chinese restaurants bring me and my mom closer to our history — and each other

Every time I walk into Swatow — a Chinatown mainstay on Spadina Avenue — one of the owners greets me with a huge smile. “Welcome back!” he says to me in Cantonese, looking behind me. “But where’s your mom?” That friendly face, Guang Bai, also known to many as the man with the photogenic memory, has logged thousands of customers’ faces and names over 42 years of running the restaurant. One of those faces being my mom, who first entered those doors as a student at the University of Toronto. “It was one of the only places I found that cured my homesickness,” she told me. “Canada was incredibly lonely, but a taste of their noodle soup immediately put tears in my eyes.” Four decades after opening, the eatery continues to be a bustling hub in the city, serving up more than 200 items (not including off-menu items). Whether for a quick and affordable lunch of wonton soup under \$10 or for late-night

09:02 i-Code: An Integrative and Composable Multimodal Learning Framework. (arXiv:2205.01818v1 [cs.LG])

Human intelligence is multimodal; we integrate visual, linguistic, and acoustic signals to maintain a holistic worldview. Most current pretraining methods, however, are limited to one or two modalities. We present i-Code, a self-supervised pretraining framework where users may flexibly combine the modalities of vision, speech, and language into unified and general-purpose vector representations. In this framework, data from each modality are first given to pretrained single-modality encoders. The encoder outputs are then integrated with a multimodal fusion network, which uses novel attention mechanisms and other architectural innovations to effectively combine information from the different modalities. The entire system is pretrained end-to-end with new objectives including masked modality unit modeling and cross-modality contrastive learning. Unlike previous research using only video for pretraining, the i-Code framework can dynamically process single, dual, and triple-modality data

09:02 Pre-RTL DNN Hardware Evaluator With Fused Layer Support. (arXiv:2205.01729v1 [cs.AR])

With the popularity of the deep neural network (DNN), hardware accelerators are demanded for real time execution. However, lengthy design process and fast evolving DNN models make hardware evaluation hard to meet the time to market need. This paper proposes a pre-RTL DNN hardware evaluator that supports conventional layer-by-layer processing as well as the fused layer processing for low external bandwidth requirement. The evaluator supports two state-of-the-art accelerator architectures and finds the best hardware and layer fusion group The experimental results show the layer fusion scheme can achieve 55.6% memory bandwidth reduction, 36.7% latency improvement and 49.2% energy reduction compared with layer-by-layer operation.

04.05.2022
04:22 A Real Time 1280x720 Object Detection Chip With 585MB/s Memory Traffic. (arXiv:2205.01571v1 [cs.AR])

Memory bandwidth has become the real-time bottleneck of current deep learning accelerators (DLA), particularly for high definition (HD) object detection. Under resource constraints, this paper proposes a low memory traffic DLA chip with joint hardware and software optimization. To maximize hardware utilization under memory bandwidth, we morph and fuse the object detection model into a group fusion-ready model to reduce intermediate data access. This reduces the YOLOv2's feature memory traffic from 2.9 GB/s to 0.15 GB/s. To support group fusion, our previous DLA based hardware employes a unified buffer with write-masking for simple layer-by-layer processing in a fusion group. When compared to our previous DLA with the same PE numbers, the chip implemented in a TSMC 40nm process supports 1280x720@30FPS object detection and consumes 7.9X less external DRAM access energy, from 2607 mJ to 327.6 mJ.

03.05.2022
07:02 Brainish: Formalizing A Multimodal Language for Intelligence and Consciousness. (arXiv:2205.00001v1 [cs.AI])

Having a rich multimodal inner language is an important component of human intelligence that enables several necessary core cognitive functions such as multimodal prediction, translation, and generation. Building upon the Conscious Turing Machine (CTM), a machine model for consciousness as proposed by Blum and Blum (2021), we describe the desiderata of a multimodal language called Brainish, comprising words, images, audio, and sensations combined in representations that the CTM's processors use to communicate with each other. We define the syntax and semantics of Brainish before operationalizing this language through the lens of multimodal artificial intelligence, a vibrant research area studying the computational tools necessary for processing and relating information from heterogeneous signals. Our general framework for learning Brainish involves designing (1) unimodal encoders to segment and represent unimodal data, (2) a coordinated representation space that relates and composes

02.05.2022
19:10 Researchers design simpler magnets for twisty facilities that could lead to steady-state fusion operation

Scientists have used a mathematical technique to design powerful magnets with straighter shapes for stellarator fusion facilities, allowing for easier manufacturing and maintenance.

09:42 Runaway electron velocity-space observation regions of bremsstrahlung hard X-ray spectroscopy. (arXiv:2204.14140v1 [physics.plasm-ph])

The reconstruction of the distribution function of runaway electrons (RE) in magnetically confined fusion plasmas gives insights on the runaway electron beam dynamics during plasma disruptions. In view of enabling a two-dimensional, energy-pitch reconstruction of the RE velocity space, in this work we present a calculation of the weight functions for the bremsstrahlung emission by the REs. The weight functions allow bridging the bremsstrahlung spectrum with the RE velocity space, as they tell the region of the velocity space that contributes to a particular spectral measurement. The results are applied to investigate the RE velocity-space sensitivity of the hard X-ray diagnostic installed at the Joint European Torus.

29.04.2022
11:23 Machine learning, harnessed to extreme computing, aids fusion energy development

Linking techniques from machine learning with advanced numerical simulations, MIT researchers take an important step in state-of-the-art predictions

09:33 Machine learning, harnessed to extreme computing, aids fusion energy development

MIT research scientists Pablo Rodriguez-Fernandez and Nathan Howard have completed one of the most demanding calculations in fusion

28.04.2022
20:43 Researchers design simpler magnets for twisty facilities that could lead to steady-state fusion operation

Harnessing the power that makes the sun and stars shine could be made easier by powerful magnets with straighter shapes than have been made before. Researchers linked to the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL) have found a way to create such magnets for fusion facilities known as stellarators.

First← Previous12345678910Previous →