[en] The real-time amplitude estimation of selective harmonics from an Avalanche Photo Diode (APD) signal of a Motion Stark Effect diagnostic is addressed using a Kalman filter. The proposed technique is shown to be much more robust and provide less noisy estimates than a lock-in amplifier scheme. In addition, the negative impact of Edge Localised Modes (ELMs) is minimized, reducing significantly the biasing in the amplitude estimation and ultimately allowing for the pitch angle estimation in the vicinity of the ELM. The inherent biasing in the amplitude estimation due to the 50Hz modulation in the NBI power grid is also easily circumvented with such a technique, rendering dispensable any further filtering of the data

[en] During the first Large Hadron Collider (LHC) shutdown period, software for the LHC Beam-based Feedback Controller (BFC) and accompanying Service Unit (BFSU) was migrated to new 64-bit multi-core hardware and to a new version of CERN's FESA realtime framework. This coincided with the transfer of responsibility to a new software team, charged with readying the systems for beam in 2015 as well as maintaining and improving the code-base in the future. In 90'000+ existing lines of code, a new testing framework was developed which would not only serve to define the system's functional specification, but also provide acceptance tests for future releases. This paper presents how the BFC and BFSU systems were decoupled from each other as well as from the LHC plant's measurement and actuator systems, thus allowing simulation-data driven instances to be deployed in a test environment. It also describes the resulting Java-based Domain-Specific Language (DSL) which allows the formation of repeatable acceptance tests. (author)

[en] Magnetic and turbulent activity in a tokamak plasma discharge may evolve in a complex trend during the discharge. In the ISTTOK tokamak, characterised by a short 30ms pulse duration, particle confinement of the order 0.3 ms and typical burst-like activity, the analysis of the time scales involved in the plasma activity is even more demanding. In this work, use is made of the Empirical Mode Decomposition (EMD) method as a tool for turbulence and MHD instability analysis obtained with Langmuir and magnetic probes, respectively. The time evolution of the energy content and wavenumber characterisation of fluctuating potential, edge plasma density and perturbed magnetic fields is investigated, with special emphasis on discharges where both limiter and electrode biasing were used to induce local sheared electric fields, thus affecting turbulence and consequently particle confinement

[en] The understanding of the nature, driving mechanisms, and dynamics of turbulence activity in tokamak plasmas is a major challenge in fusion research. In the ISTTOK tokamak, edge turbulence activity has been extensively studied using both limiter and electrode biasing to induce local sheared electric fields, thus affecting turbulence and consequently particle confinement. Due to the fast time scales involved plasmas (pulse length of 30 ms and particle confinement of the order of 0.3 ms) and the turbulent nature of ISTTOK plasmas, not only turbulence but also magnetohydrodynamic (MHD) activity occurs at rapid bursts. Therefore, resolving the modes both spatially and during time (in particular, regarding the energy/frequency spectrum) requires advanced time-frequency analysis tools, capable of tracking short lasting (0.1 ms) MHD modes. In this article, we explore the capabilities of the empirical mode decomposition method as a tool for turbulence and MHD instability analysis. The time evolution of the energy content associated with fluctuating potential, edge plasma density (obtained from Langmuir probes), and perturbed magnetic fields (Mirnov coil signals) is obtained, and a comparison is made with other advanced time-frequency signal estimators

[en] The increasing importance of real-time measurements and control systems in JET experiments, regarding e.g. Internal Transport Barrier (ITB) and q-profile control, has motivated the development of a real-time motional Stark effect (MSE) system. The MSE diagnostic allows the measurement of local magnetic fields in different locations along the neutral beam path providing, therefore, local measurement of the current and q-profiles. Recently in JET, an upgrade of the MSE diagnostic has been implemented, incorporating a totally new system which allows the use of this diagnostic as a real-time control tool as well as an extended data source for off-line analysis. This paper will briefly describe the technical features of the real-time diagnostic with main focus on the system architecture, which consists of a VME crate hosting three PowerPC processor boards and a fast ADC, all connected via Front Panel Data Port (FPDP). The DSP algorithm implements a lockin-amplifier required to demodulate the JET MSE signals. Some applications for the system will be covered such as: feeding the real-time equilibrium reconstruction code (EQUINOX) and allowing the full coverage analysis of the Neutral Beam time window. A brief comparison between the real-time MSE analysis and the off-line analysis will also be presented

[en] The successful control of a fusion plasma discharge relies on the continuous improvement of existing diagnostics and on the development of new ones based on innovative concepts. It is equally important to provide efficient numerical tools capable of extracting, in real time, the most important physical quantities emerging from the collected data. This encompasses, for instance, the determination of the wavenumber and amplitude tracking of plasma instabilities, a necessary ingredient for establishing in real time the type of feedback control stabilization scheme to be used. In this paper, the characterization of the wavenumber spectra of unstable modes observed in ISTTOK plasma discharges is addressed using a set of two poloidal arrays of magnetic coils (Mirnov coils). In particular, the prospect for the estimation of the poloidal wavenumber in real time using recursive filtering techniques is investigated. It is shown that, in contrast with typical cross-spectrum analysis, where a large dataset is required to estimate the wavenumbers, with this innovative technique, the wavenumber spectra may be obtained in faster time scales.

[en] The real-time tracking of instantaneous quantities such as frequency, amplitude, and phase of components immerse in noisy signals has been a common problem in many scientific and engineering fields such as power systems and delivery, telecommunications, and acoustics for the past decades. In magnetically confined fusion research, extracting this sort of information from magnetic signals can be of valuable assistance in, for instance, feedback control of detrimental magnetohydrodynamic modes and disruption avoidance mechanisms by monitoring instability growth or anticipating mode-locking events. This work is focused on nonlinear Kalman filter based methods for tackling this problem. Similar methods have already proven their merits and have been successfully employed in this scientific domain in applications such as amplitude demodulation for the motional Stark effect diagnostic. In the course of this work, three approaches are described, compared, and discussed using magnetic signals from the Joint European Torus tokamak plasma discharges for benchmarking purposes

[en] Magnetohydrodynamic (MHD) activity in magnetically confined fusion experiments is often associated with detrimental effects such as increased radial transport and consequent loss of confinement. In particular, the (2,1) neoclassical tearing mode (NTM), when proceeding to mode-locking, is a potentially disruptive instability hence with the potential to compromise the mechanical integrity of the machine. It is therefore quite significant to be able to characterize in real-time the most virulent and performance limiting instabilities such that adequate mitigation or complete stabilization using feedback control methods are employed during the plasma discharge. This work proposes a Kalman filter (KF) based mechanism for providing, in real-time, the amplitude and phase evolution of instabilities within a predefined set of mode numbers. The method relies on two KF implementations: a non-linear KF isolating the non-stationary dominant signal component of a sensor measurement and subsequently a linear KF which projects the former, for a collection of sensors, onto a predefined set of mode numbers. A basic overview of algorithms commonly used for real-time mode number analysis is also presented along with applications of the proposed algorithm to recently recorded data of the Joint European Torus (JET) tokamak. (paper)

[en] The heavy ion beam probe (HIBP) diagnostic permits the measurements of electron density and plasma potential as well as their fluctuations. The use of a multiple cell array detector (MCAD) spreads the HIBP capabilities to make simultaneous measurements across the plasma column. Two sets of current-to-voltage converters with respective bandwidths of 400 kHz and 4 MHz have been tested. This upgrade of the MCAD electronics has allowed the observation of quasicoherent fluctuations in bulk plasma at frequencies between 100 and 200 kHz. The localization and frequency of bulk plasma fluctuations correlate with the magnetohydranamic activity of the plasma.

[en] We have detected 90 objects with periods and light-curve structures similar to those of field δ Scuti stars using the Massive Compact Halo Object (MACHO) Project database of Galactic bulge photometry. If we assume similar extinction values for all candidates and absolute magnitudes similar to those of other field high-amplitude δ Scuti stars (HADS), the majority of these objects lie in or near the Galactic bulge. At least two of these objects are likely foreground δ Scuti stars, one of which may be an evolved nonradial pulsator, similar to other evolved, disk-population δ Scuti stars. We have analyzed the light curves of these objects and find that they are similar to the light curves of field δ Scuti stars and the δ Scuti stars found by the Optical Gravitational Lens Experiment (OGLE). However, the amplitude distribution of these sources lies between those of low- and high-amplitude δ Scuti stars, which suggests that they may be an intermediate population. We have found nine double-mode HADS with frequency ratios ranging from 0.75 to 0.79, four probable double- and multiple-mode objects, and another four objects with marginal detections of secondary modes. The low frequencies (5-14 cycles day-1) and the observed period ratios of ∼0.77 suggest that the majority of these objects are evolved stars pulsating in fundamental or first overtone radial modes. (c) 2000 The American Astronomical Society