Tokamak Control Research Articles

MHD stability constraints on divertor heat flux width in DIII-D

Abstract The radial width of the exhaust heat flux flowing in the SOL of DIII-D is found to expand at high input power and plasma density, consistent with MHD ballooning stability limits. At low heating power, ~3 MW, the SOL width remains constant and consistent with established empirical scaling dependent only on the midplane poloidal field. At high heating power, ~ 13 MW a higher separatrix density, and resulting higher separatrix pressure is required for divertor detachment. The separatrix pressure gradient at the separatrix continues to increase with density until it begins to saturate at levels ~50% above the calculated ideal MHD ballooning limit. Examination of the separate contributions to the pressure gradient from electrons and ions reveals the ion pressure gradient to saturate more strongly than the electron pressure gradient. Potential analysis issues leading to the measured pressure gradient exceeding the ballooning limit are discussed. At high density, particularly for detached divertor plasmas, the SOL width for temperature and density expand modestly, ~30–50%. The divertor plasma density profile in detachment also reflects this trend, expanding in the radial direction a factor of 2–3. Despite the SOL width expansion at the highest power and density no degradation of the pedestal and resulting core confinement is observed with the additional density at high power. These results imply a more favorable scaling for divertor heat flux control in future reactor-scale tokamaks than predicted by existing empirical scaling.

Weighted majority rule ensemble classifier for sensor fault classification for plasma position control in Tokamaks

Tokamaks are the most promising devices which employ magnetic confinement of Deuterium and Tritium plasma to achieve nuclear fusion for power generation. Magnetic flux sensors are employed to measure the position of the plasma column inside the device. Faults occurring in these sensors may cause the failure of control of the plasma position, thereby terminating the fusion process. In this paper, we present a comparison of different Machine Learning (ML) algorithms to detect and classify the sensors’ faults. We applied four different ML algorithms; namely, Logistic Regression based Multi-layer Perceptron Classifier (MLP), Support Vector Classifier (SVC), K-Nearest Neighbor (KNN), and Decision Tree (DT) as the individual ML classifiers for the task of sensor fault classification. We propose an ensemble classifier (ECF) using the classifiers, as mentioned earlier. We provide a comparative assessment of the classification accuracy of different classifiers, namely, MLP, SVC, KNN, DT, and ECF. The ECF utilized a weighted majority voting method to combine the decision of individual faults classifiers. Overall, robustness to the miss-classification of new data by the individual sensor fault classifiers by applying the ECF is reported.

Measurement of Time Delay between Magnetic Diagnostic Coils in Tokamak Using Cross-Spectral Analysis

Delay in detecting the position and boundaries of the plasma confined by a tokamak increases plasma instability and decreases the accuracy of plasma control in tokamak. In this article, two techniques called ‘‘periodogram’’ and ‘‘cross-spectral analysis’’ are used to provide a quantitative approach for computing the time delay between data received by two diversity coils from one source (plasma fluctuations) in IR-T1 tokamak. To analysis the achieved data after performing the Fourier transform on the data, the periodogram, square-coherency and spectrum phase diagram of the data were plotted in terms of frequency at three different time intervals. In the time interval, $$29000<t<3000$$ $$\mu$$ s, when the plasma current is almost constant, the average phase shift between the $$\omega$$ -frequencies components of the data was obtained by using the phase-spectrum analysis. At the frequency of the highest coherency the time delay ( $$\tau_{d}=1.831$$ $$\mu$$ s) between the data of the Rogowski and Mirnov coils was calculated by considering the average phase shift between them.

Fast modeling of turbulent transport in fusion plasmas using neural networks

We present an ultrafast neural network model, QLKNN, which predicts core tokamak transport heat and particle fluxes. QLKNN is a surrogate model based on a database of 3 × 108 flux calculations of the quasilinear gyrokinetic transport model, QuaLiKiz. The database covers a wide range of realistic tokamak core parameters. Physical features such as the existence of a critical gradient for the onset of turbulent transport were integrated into the neural network training methodology. We have coupled QLKNN to the tokamak modeling framework JINTRAC and rapid control-oriented tokamak transport solver RAPTOR. The coupled frameworks are demonstrated and validated through application to three JET shots covering a representative spread of H-mode operating space, predicting the turbulent transport of energy and particles in the plasma core. JINTRAC–QLKNN and RAPTOR–QLKNN are able to accurately reproduce JINTRAC–QuaLiKiz Ti,e and ne profiles, but 3–5 orders of magnitude faster. Simulations which take hours are reduced down to only a few tens of seconds. The discrepancy in the final source-driven predicted profiles between QLKNN and QuaLiKiz is on the order of 1%–15%. Also the dynamic behavior was well captured by QLKNN, with differences of only 4%–10% compared to JINTRAC–QuaLiKiz observed at mid-radius, for a study of density buildup following the L–H transition. Deployment of neural network surrogate models in multi-physics integrated tokamak modeling is a promising route toward enabling accurate and fast tokamak scenario optimization, uncertainty quantification, and control applications.

Data-driven disruption prediction in GOLEM Tokamak using ensemble classifiers

A robust disruption prediction system is mandatory in a Tokamak control system as the disruption can cause malfunctioning of the plasma-facing components and impair irrecoverable structural damage to the vessel. To mitigate the disruption, in this article, a data-driven based disruption predictor is developed using an ensemble technique. The ensemble algorithm classifies disruptive and non-disruptive discharges in the GOLEM Tokamak system. Ensemble classifiers combine the predictive capacity of several weak learners to produce a single predictive model and are utilized both in supervised and unsupervised learning. The resulting final model reduces the bias, minimizes variance and is unlikely to over-fit when compared to the individual model from a single algorithm. In this paper, popular ensemble techniques such as Bagging, Boosting, Voting, and Stacking are employed on the time-series Tokamak dataset, which consists of 117 normal and 70 disruptive shots. Stacking ensemble with REPTree (Reduced Error Pruning Tree) as a base learner and Multi-response Linear Regression as meta learner produced better results in comparison to other ensembles. A comparison with the widely employed stand-alone machine learning algorithms and ensemble algorithms are illustrated. The results show the excellent performance of the Stacking model with an F1 score of 0.973. The developed predictive model would be capable of warning the human operator with feedback about the feature(s) causing the disruption.

Real-time detection and correction of frequency sweeping non-linearities of FMCW reflectometry microwave sources

FMCW reflectometry density profile measurements will be used to complement or replace magnetic measurements for plasma position control in future plasma fusion reactors, such as ITER and DEMO. Although FMCW reflectometry is a well established diagnostic technique, on present experimental devices with short discharges (typically of a few seconds), on-line measurement quality validation and correction is not a common practice. One of the technique's main requirements is to ensure that the microwave sources always produce a linear frequency sweep within the probing frequency range. Failing to do so leads to inaccurate density profile reconstruction and, hence, to erroneous plasma position feedback to the tokamak control loops. Traditionally, static calibrations are used to build the voltage ramps driving the reflectometer's VCO (Voltage Controlled Oscillator) sources. By using extra signals like frequency markers or calibrated delay lines, the experimental reflectometry measurements can be further calibrated, off-line, to minimize frequency sweep linearity related distortions. We herein propose a technique to monitor and re-calibrate in real-time the linearity of the frequency sweeps of FMCW reflectometers. It can be applied in the long duration discharges of future devices either to correct thermal or electronic drifts of the probing frequency sources or to generate alarm signals when measurement anomalies or fundamental diagnostic malfunctions are detected. Experimental data is presented showing the benefits of the proposed technique.

Toroidal modeling of resistive wall mode stability and control in HL-2M tokamak

Effects of toroidal plasma flow, magnetic drift kinetic damping as well as feedback control, on the resistive wall mode instability in HL-2M tokamak are numerically investigated, using the linear stability codes MARS-F/K (Liu et al 2000 Phys. Plasmas 7 3681, Liu et al 2008 Phys. Plasmas 15 112503). It is found that the precession drift resonance damping due to trapped thermal particles ensures a robust passive stabilization of the n = 1 (n is the toroidal mode number) RWM in the 2 MA double-null advanced plasma scenario designed for HL-2M, provided that the toroidal flow speed is not too fast: . With two rows of magnetic control coils designed for HL-2M, the optimal poloidal location for the RWM stabilization is found to be . Toroidal modeling also shows that the plasma flow damping, drift kinetic damping and magnetic feedback can be arranged to synergistically stabilize the RWM in HL-2M, by tuning the feedback gain phase and/or including derivative actions in the control loop. The numerical results obtained by MARS-F/K are qualitatively well re-produced by an analytic single-pole model.

FPGA-based novel real-time evaluation and data quality monitoring system for tokamak high-performance GEM soft X-ray diagnostic

Based on literature review data quality monitoring topic is often being simplified in case of tokamak diagnostic systems. In previous publications Authors presented a novel model of real-time evaluation and data quality monitoring (EDQM) sub-system for use in physics experiments, mostly related to tokamak plasma measurements. The implementation of model is done in scope of monitoring raw output signals from Gas Electron Multiplier detectors used in FPGA-based soft X-ray diagnostics. System is designed to work in the real-time feedback loop for parameters control of tokamak in scope of achieving long-term plasma stability. In the paper are described results of application the DQM engine from measurements with 112-channel GEM detector and real radiation sources (isotopes, X-ray tube). The discussion provided in article covers the usefulness of applied novel EDQM techniques: technical and statistical, in scope of detection of new, corrupted output signals from the GEM detector when working under hard environment conditions. The presented in the paper results indicates the ability of the detection of subtle system defects (e.g. detector, electronics), resulting in the significant corruption of the output products of measured radiation: energy and topological spectra. This is especially important for the systems working in the real-time feedback loop. The novel approach of detection new malformed signals with ability to automatically discard those phenomena from further analysis is also described. In the summary there are described benefits of usage of the proposed novel DQM approach with indication of another unique work for further CPU extensions of the EDQM sub-system providing complex FPGA-CPU high-performance data quality monitoring unit. The EDQM system is planned to be installed in GEM SXR diagnostics on WEST tokamak.

Plasma Control in Tokamaks. Part 2.

Various systems of magnetic control of plasma position, current, and shape are considered in vertically elongated tokamaks including spherical tokamaks being in operation. The systems of Resistive Wall Modes suppression are described. Tokamaks constructions and cross-sections, structural schemes of plasma control systems are given; heed is paid to operating principles of plasma control systems; experimental results of plasma control in tokamaks are given. Various engineering realizations of magnetic plasma control systems in tokamaks are presented.

Plasma Magnetic Control in Tokamak Devices

In tokamak experimental reactors, the magnetic control system is one of the main plasma control systems that is required, together with the density control, since the very beginning, even before first operations. Indeed, the magnetic control drives the current in the external poloidal circuits in order to first achieve the breakdown conditions and, after plasma formation, to track the desired plasma current, shape and position. Furthermore, when the plasma poloidal cross-section is vertically elongated, the magnetic control takes also care of the vertical stabilization of the plasma column, and therefore it is an essential system for operation. This chapter introduces a reference architecture for plasma magnetic control in tokamaks. Given the proposed architecture, the techniques to design all the required control algorithms is also presented. Experimental results obtained on the JET and EAST tokamaks and simulations for machines currently under construction are shown to prove the effectiveness of the proposed architecture and control algorithms.

High-Speed Visible Image Acquisition and Processing System for Plasma Shape and Position Control of EAST Tokamak

Currently, the fast evolution of cameras in recent years has made them promising tools for diagnostics of Tokamak. The solution presented in this paper consists of a prototype of high-speed visible image acquisition and processing system (HVIAPs) dedicated for experimental advanced superconducting tokamak shape and position control. Graphics processing unit (GPU) and field-programmable gate array (FPGA) are typically used as accelerators or co-processors in addition to a CPU. Such a heterogeneous computing system can combine the advantages of its individual components. The main imaging equipment components used for this system are four high-resolution fast cameras equipped with fiber interface. Image data from the cameras are received by the frame grabber card based on the FPGA and transmitted to the GPU via the peripheral component interconnect express interface. The software support for the system includes low-level drivers and application programming interface libraries for all components and the algorithm developed for image processing. The offline image process results are compared to equilibrium fitting code, which is a commonly used reconstruction method, with an average error of 1.5 cm. The total processing time for one frame is less than 0.3 ms.

Actuator allocation for integrated control in tokamaks: architectural design and a mixed-integer programming algorithm

Plasma control systems (PCS) in tokamaks need to fulfill a number of control tasks to achieve the desired physics goals. In present-day devices, actuators are usually assigned to a single control task. However, in future tokamaks, only a limited set of actuators is available for multiple control tasks at the same time. The priority to perform specific control tasks may change in real-time due to unforeseen plasma events and actuator availability may change due to failure. This requires the real-time allocation of available actuators to realize the requests by the control tasks, also known as actuator management.In this paper, we analyze possible architectures to interface the control tasks with the allocation of actuators inside the PCS. Additionally, we present an efficient actuator allocation algorithm for Heating and Current Drive (H&CD) actuators. The actuator allocation problem is formulated as a Mixed-Integer Quadratic Programming optimization problem, allowing to quickly search for the best allocation option without the need to compute all allocation options. The algorithms performance is demonstrated in examples involving the full proposed ITER H&CD system, where the desired allocation behavior is successfully achieved. This work contributes to establishing integrated control routines with shared actuators on existing and future tokamaks.

Hierarchical robust systems for magnetic plasma control in tokamaks with adaptation

We consider synthesis and modeling of hierarchical robust magnetic control systems for plasma in tokamaks with adaptation of the magnetic axis' vertical position. The studies were done for ITER (Cadarache, France) and Globus-M (St. Petersburg, Russia) tokamaks. Our original synthesis method ensures an improvement in performance for the shape of the plasma by resolving the contradiction between position and plasma shape via adaptation in a hierarchical system. Numerical modeling of synthesized control systems on multivariable linear models of plasma and nonlinear plasma-physics code DINA that unites basic equations of plasma dynamics in a magnetic field has shown the efficiency of the proposed approach and a possibility to apply it in a physical experiment.

The control system of the 1-MW TCV tokamak heating neutral beam injector

The control system of the neutral beam injector is presented that was designed and constructed at the Budker Institute of Nuclear Physics in Novosibirsk, for the TCV tokamak in Lausanne, Switzerland. The system hardware includes an industrial computer with reconfigurable PCI Express input–output boards, as well as interfacing electronics, which ensures galvanic isolation and electrical compatibility with the injector units. The software and firmware that implement the control system algorithm by means of both the industrial computer and the FPGAs of the reconfigurable boards are written using LabVIEW. The injection parameters are calculated using the beam-power model; the graphical representation and storage of the results are carried out in MATLAB. A communication interface with the tokamak control system is also provided for data exchange and synchronization of operation.

Improvements of magnetic measurements for plasma control in KSTAR tokamak

The improvements in the magnetic measurement are carried out to provide more accurate plasma parameters required for better plasma control in the KSTAR tokamak. Firstly, the large noise width in the magnetic diagnostic (MD) sensor signals (up to δV/V=∼20%), which caused from the IVCC power supply being essential for sustaining high elongated plasmas, can be reduced down to ∼1% by using commercial anti-aliasing (AA) electronic filters with the cut-off frequency fc of 0.86kHz. Secondly, the time response of ∼1.0ms required for the fast vertical control is achieved by using new AA filter (fc=2.2kHz) which has been developed to minimize the time delay due to the previous filter. In this paper, the characteristics of the AA filters are described, and the improvements in the magnetic measurements are presented.

A current-pulsed power supply with rapid rising and falling edges for magnetic perturbation coils on the J-TEXT tokamak

This study presents the design and principle of a current-pulsed power supply (CPPS) for the tearing mode (TM) feedback control of the J-TEXT tokamak. CPPS is a new method of stabilizing large magnetic islands and accelerating mode rotation through the use of modulated magnetic perturbation. In this application, continuous magnetic perturbation pulse trains with frequency of 1kHz to kHz, amplitude of 0.25G, and duty ratio of 20%–50% are required generating via in-vessel magnetic coils. A modified topology based on buck chopper is raised to satisfy the demands of inductive load. This modified topology is characterized by high frequency, rapid rising and falling edges, and large amplitude of current pulses. Appropriate RCD snubber circuit is applied to protect the Insulated Gate Bipolar Transistor (IGBT) switch device. Equipment with peak current that reaches 1kA, frequency that ranges from 1kHz to 3kHz, and rising and falling time within 100μs was constructed and applied to physical experiment.

Soft X-ray tomography in support of impurity control in tokamaks

This contribution reviews an important example of current developments in diagnostic systems and data analysis tools aimed at improved understanding and control of transport processes in magnetically confined high temperature plasmas. The choice of tungsten for the plasma facing components of ITER and probably also DEMO means that impurity control in fusion plasmas is now a crucial challenge. Soft X-ray (SXR) diagnostic systems serve as a key sensor for experimental studies of plasma impurity transport with a clear prospective of its control via actuators based mainly on plasma heating systems. The SXR diagnostic systems typically feature high temporal resolution but limited spatial resolution due to access restrictions. In order to reconstruct the spatial distribution of the SXR radiation from line integrated measurements, appropriate tomographic methods have been developed and validated, while novel numerical methods relevant for real-time control have been proposed. Furthermore, in order to identify the main contributors to the SXR plasma radiation, at least partial control over the spectral sensitivity range of the detectors would be beneficial, which motivates for developments of novel SXR diagnostic methods. Last, but not least, semiconductor photosensitive elements cannot survive in harsh conditions of future fusion reactors due to radiation damage, which calls for development of radiation hard SXR detectors. Present research in this field is exemplified on recent results from tokamaks COMPASS, TORE SUPRA and the Joint European Torus JET. Further planning is outlined.

RETRACTED ARTICLE: Edge-Localized Modes (ELM) Modification with Transport Barrier Control in Tokamak by Analytical Strategies

High confinement mode (H-mode) in tokamaks is characterized by the formation of a transport barrier at the edge of the plasma. The H-mode increases the plasma energy confinement time by around a factor of two compared to the Low confinement mode (L-mode). This result made a great step towards achieving the higher temperatures and pressures needed to create ignition conditions. The Grad–Shafranov Equation describes the magnetic flux distribution of plasma in an axisymmetric system like a tokamak. In this paper, we have calculated the magnetic flux surfaces in IR-T1 tokamak by two methods. An analytical solution of magnetic surfaces approximated by equilibrium calculation of GS equation based on first three (and then four) terms of expansion in flux function. Results of two methods are in good agreement with each other.

Plasma q-profile control in tokamaks using a damping assignment passivity-based approach

The IDA-PBC based on PCH model for tokamak q-profile is investigated. Two scenarios are carried out. The first one is the resistive diffusion model for the magnetic poloidal flux. The second one is extended with the thermal diffusion. A feedforward control is used to ensure the compatibility with the actuator physical ability. An IDA-PBC feedback is proposed to improve the system stabilization and convergence speed. The controllers are validated in the simulation using RAPTOR code and tested in TCV, the result is analyzed and the followed discussion proposed the required improvement for the next experiments.

Measurement of eddy-current distribution in the vacuum vessel of the Sino-UNIted Spherical Tokamak.

Eddy currents have an important effect on tokamak plasma equilibrium and control of magneto hydrodynamic activity. The vacuum vessel of the Sino-UNIted Spherical Tokamak is separated into two hemispherical sections by a toroidal insulating barrier. Consequently, the characteristics of eddy currents are more complex than those found in a standard tokamak. Thus, it is necessary to measure and analyze the eddy-current distribution. In this study, we propose an experimental method for measuring the eddy-current distribution in a vacuum vessel. By placing a flexible printed circuit board with magnetic probes onto the external surface of the vacuum vessel to measure the magnetic field parallel to the surface and then subtracting the magnetic field generated by the vertical-field coils, the magnetic field due to the eddy current can be obtained, and its distribution can be determined. We successfully applied this method to the Sino-UNIted Spherical Tokamak, and thus, we obtained the eddy-current distribution despite the presence of the magnetic field generated by the external coils.