Search

openaire: EC/H2020/633053/EU//EUROfusion A remote LIBS system is being considered as an analysis tool in ITER for monitoring the erosion and fuel retention in the first wall. This necessitates further investigation of the performance of LIBS for thick co-deposited layers consisting of ITER-relevant materials. The main goal of this work is determining the fuel content of the samples and D depth profile by LIBS and to compare the results with those obtained by other methods. The studied samples were Be-based mixed coatings on W substrates, containing D and in some cases C, O, or both. These impurities are relevant not only for ITER, but also for other fusion devices, such as JET-ILW. The laser ablation was performed at 10 mbar Ar pressure using a 5 ns pulse Nd:YAG laser at 1064 nm. Suitable Be and Ar spectral lines were employed for the evaluation of the electron temperature of the plasma using multi-elemental Saha-Boltzmann (MESB) plots. The electron density was obtained from the Stark broadening of the Dα and Hα spectral lines following deconvolution. The average elemental content of each coating was then obtained by calibration-free LIBS (CF-LIBS) and are in agreement with other techniques (TOF–ERDA, IBA).

High speed visible imaging allows to visualize the 2-dimensional dynamics of fast phenomena in the boundary layer of fusion plasmas. Here we describe the two high speed visible cameras currently operating in the JET tokamak, enabling the simultaneous observation of a large fraction of the JET tokamak vacuum vessel with the appropriate configuration for many different plasma phenomena. As an example we discuss recent observations on the spatial and temporal dynamics of visible emission at the outer wall during Edge Localized Modes (ELMs) and the penetration of Shattered Injected Pellet (SPI) into the plasma.

openaire: EC/H2020/633053/EU//EUROfusion Estimates for 2D distributions of electron density, ne, and temperature, Te, have been obtained using reflection-corrected tomographic reconstructions of deuterium Balmer line emission in the JET divertor. The solutions are obtained by a two-dimensional Monte Carlo optimization model, based on intensity ratios of the Balmer Dα, Dγ and Dε emission and the ADAS photon emission coefficients with additional constraints provided by line-integrated spectroscopic measurements of ne and Te. Synthetic benchmarks utilizing EDGE2D-EIRENE simulation outputs yield reproduction of reference ne and Te distributions within 25% with statistical error estimates of 5–15%. Experimental analysis of an L-mode ne ramp shows extension and movement of the outer divertor high-ne region with ne,div up to 1.5 × 1020 m−3 from the outer strike point to above the X-point and decrease of the divertor Te to the range of 0.5–1.0 eV within 25% agreement with the line-integrated measurements. Comparison between the local estimates and their corresponding line integrals suggests underestimation of ne and Te by up to 75% and 50%, respectively, due to the geometric effects of line integration.

In ITER, a remote handling laser-induced breakdown spectroscopy (LIBS) system is considered to be able to observe regions where deposition of thick layers is expected to occur and assess the retention of fuel on wall structures. The H and D contents of surface layers on divertor and first-wall materials have been intensely studied by LIBS but mostly by depth profile analysis of D. As far as we know, quantitative analysis of H/D in BeW mixed layer by calibration free-LIBS (CF-LIBS) has not been performed. Thus, the aim of this work is the quantification of the relative concentrations of D and H in Be-based material samples by CF-LIBS. The samples used were Be/W(67:33) as mixed homogeneous coatings (2 μm) on Mo substrates. Laser ablation was performed using a 1064 nm laser with 5 ns pulses. In order to enhance the resolution to distinguish H, D and in future T from each other, two different pressures were used: high vacuum and 0.5 mbar of Ar. Suitable Be and W spectral lines (without interferences and self-absorption) have been selected for precise evaluation of the electron temperature, T e , of the plasma using the Boltzmann plots. The electron density, n e , was extracted from the Saha equation using the average electron temperatures obtained from the Stark broadening of the H α line. With these values, the D content of the samples has been calculated by CF LIBS as ∼4.7% ± 2.9%. These results are in a good agreement with thermal desorption spectroscopy measurements, which gives a 4–5 at% for the D content. In addition, the depth profile is similar to that recorded using secondary ion mass spectrometry.

In this paper, we consider a linear supervised dimension reduction method for classification settings: stochastic discriminant analysis (SDA). This method matches similarities between points in the projection space with those in a response space. The similarities are represented by transforming distances between points to joint probabilities using a transformation which resembles Student’s t-distribution. The matching is done by minimizing the Kullback–Leibler divergence between the two probability distributions. We compare the performance of our SDA method against several state-of-the-art methods for supervised linear dimension reduction. In our experiments, we found that the performance of the SDA method is often better and typically at least equal to the compared methods. We have made experiments with various types of data sets having low, medium, or high dimensions and quite different numbers of samples, and with both sparse and dense data sets. If there are several classes in the studied data set, the low-dimensional projections computed using our SDA method provide often higher classification accuracies than the compared methods.

The academic security research community has studied the Android malware detection problem extensively. Machine learning methods proposed in previous work typically achieve high reported detection performance on fixed datasets. Some of them also report reasonably fast prediction times. However, most of them are not suitable for real-world deployment because requirements for malware detection go beyond these figures of merit. In this paper, we introduce several important requirements for deploying Android malware detection systems in the real world. One such requirement is that candidate approaches should be tested against a stream of continuously evolving data. Such streams of evolving data represent the continuous flow of unknown file objects received for categorization, and provide more reliable and realistic estimate of detection performance once deployed in a production environment. As a case study we designed and implemented an ensemble approach for automatic Android malware detection that meets the real-world requirements we identified. Atomic Naive Bayes classifiers used as inputs for the Support Vector Machine ensemble are based on different APK feature categories, providing fast speed and additional reliability against the attackers due to diversification. Our case study with several malware families showed that different families are detected by different atomic classifiers. To the best of our knowledge, our work contains the first publicly available results generated against evolving data streams of nearly 1 million samples with a model trained over a massive sample set of 120,000 samples.

In this article we examine the methods for detecting and predicting aging related process sensor failures by analyzing the noise of the sensor output signal. The study uses data from non-differential and differential pressure transmitters used in the pressure and water level measurements of the reactor pressure vessels of units 1 and 2 of the Olkiluoto nuclear power plant in Finland. The article contains a review of the current methods for detection of sensor failures. Additionally, we present a new method for detecting changes in the sensor output signal. The method creates fingerprints of the power spectra of the sensors by using Principal Component Analysis (PCA). The changes in these fingerprints together with the measurements of the redundant sensors can be used to detect indications of some of the impending sensor failures. In the experimental study we are able to produce stable fingerprints for both the non-differential and differential pressure transmitters. Also, a potential failure in one of the differential pressure transmitters in Olkiluoto unit 2 is detected by inspecting the fingerprints and analyzing the spectral changes of the transmitter output signal.

The discrimination of cloudy from cloud-free pixels is required in almost any estimate of a parameter retrieved from satellite data in the ultraviolet (UV), visible (VIS) or infrared (IR) parts of the electromagnetic spectrum. In this paper we report on the development of a neural network (NN) algorithm to estimate cloud fractions using radiances measured at the top of the atmosphere with the NASA-Aura Ozone Monitoring Instrument (OMI). We present and discuss the results obtained from the application of two different types of neural networks, i.e., extreme learning machine (ELM) and back propagation (BP). The NNs were trained with an OMI data sets existing of six orbits, tested with three other orbits and validated with another two orbits. The results were evaluated by comparison with cloud fractions available from the MODerate Resolution Imaging Spectrometer (MODIS) flying on Aqua in the same constellation as OMI, i.e., with minimal time difference between the OMI and MODIS observations. The results from the ELM and BP NNs are compared. They both deliver cloud fraction estimates in a fast and automated way, and they both performs generally well in the validation. However, over highly reflective surfaces, such as desert, or in the presence of dust layers in the atmosphere, the cloud fractions are not well predicted by the neural network. Over ocean the two NNs work equally well, but over land ELM performs better.