|
This article describes the feedback control system in magnetic confinement devices, aiming to mitigate magnetohydrodynamic (MHD) instabilities and optimize plasma confinement within the Keda Torus eXperiment (KTX). The system includes saddle sensors as the input signal, a real-time control system as the core controller, a digital power amplifier (DPA) array as the actuator, and saddle coils as the brake to restrain the MHD instabilities. A flexible, distributed real-time control system based on field-programmable gate array (FPGA) has been developed and implemented for MHD control in KTX. This is the first time FPGA has been used to verify presupposed magnetic field mode, perform data acquisition, real-time calculation, and real-time feedback in KTX. It fulfills the real-time requirements of the experiment completely. The system is presently utilized in MHD control experiments, successfully achieving experiments at the current stage that reach the limit of Ohmic discharge, with good reproducibility. Introduction Magnetic confinement encompasses various configurations for achieving controlled fusion reactions, including the well-known tokamak and stellarator designs [1]. In addition to these configurations, the reversed field pinch (RFP) stands as an intriguing alternative. Unlike the tokamak, which relies on intricate external magnetic field control, the RFP device primarily relies on the self-organization of the magnetic field [2]. While the RFP configuration offers unique advantages, such as simplified magnetic field geometry, it also faces challenges related to plasma instabilities that can degrade plasma confinement [3]. Among these instabilities, the ones of particular concern in RFP devices are the MHD instabilities, such as resistive wall modes (RWMs) and tearing modes (TMs), characterized by longer spatial structures and larger wavelengths. These instabilities can significantly impact plasma confinement and even lead to premature termination of plasma discharges. Thus, effectively managing these MHD instabilities through feedback control becomes crucial for achieving improved performance in RFP devices [4], [5]. Several RFP devices, including RFX-mod, EXTRAP T2R, and RELAX, have successfully integrated advanced feedback control, achieving remarkable progress in numerous experiments [6], [7], [8]. In particular, in RFX-mod, significant improvement in parameters of plasma discharge and plasma confinement has been achieved through feedback control, leading to the discovery of the quasi-single helical state that significantly enhances the performance of plasma confinement [7], [9], [10]. KTX located at the University of Science and Technology of China (USTC) is a newer member of the RFP community. KTX can be operated in either RFP or tokamak modes and is distinguished by its compact size, featuring a major radius (R) of 1.4 m and a minor radius (r) of 0.4 m. It primarily aims to enhance the performance of existing international RFP facilities [11], [12]. Table 1 provides a list of parameters between the KTX device and the aforementioned RFP devices. Although MHD feedback control system was contemplated during the initial design phase of KTX, the real-time control system, which serves as the backbone of feedback control, was not integrated initially. Thus, a primary task presented in this paper is to design and implement a real-time control system that fulfills the feedback control requirements of the KTX. During the initial phase of KTX, a data acquisition (DAQ) system critical for plasma diagnostics and feedback control was integrated. This system, designed for electromagnetic diagnostics, features over 1400 channels and utilizes saddle coils for feedback control. Built on the PCI eXtensions for Instrumentation (PXI) platform, the DAQ system employs the commercial ADLINK PXI-62022 board [13], boasting a sampling rate of 250 kSamples/s, a 16-bit resolution, and an input range of ± 10 V. However, this DAQ system does not align with the needs of real-time MHD control. As a result, we have developed an innovative real-time control system that simultaneously supports data acquisition and real-time computation [14]. Complementing this development, the Southwest Institute of Physics has designed a new DPAs system, which, together with the existing saddle coils, serves as an actuator in feedback control. While most magnetic confinement devices typically rely on general-purpose X86 processors (GPPs) for real-time control systems [15], [16], [17], FPGAs offer significant advantages. FPGAs are capable of executing parallel operations, often delivering performance levels that are an order of magnitude or more superior to GPPs [18]. Nevertheless, this performance boost comes at the cost of increased complexity in addressing timing logic issues and elevated development challenges. Recent trends, however, indicate a shift towards allocating calculations with higher timing requirements to FPGAs, while calculations without stringent timing requirements continue to be executed on CPUs. This approach leverages the speed of FPGAs while retaining the portability benefits of GPPs [19], [20]. With the objective of achieving shorter feedback cycles in the real-time feedback system, we have adopted FPGA as the computing core. Thus, this article presents an FPGA-based real-time distributed feedback control system. This article is structured as follows: Section 2 outlines the architecture of the feedback control system, encompassing the sensors and actuators, and the design of the distributed FPGA-based real-time control system. Section 3 introduces the control scheme for real-time control by FPGA. Section 4 details tests conducted to verify the effectiveness of magnetic field mode control. Lastly, Section 5 presents the experimental outcomes of real-time feedback control under tokamak discharge mode in KTX. Section snippets System overview The KTX feedback control system mainly consists of several essential components: the composite shell (CS), electromagnetic probe (EMP) array, real-time control system, DPAs, and saddle coils, as depicted in Fig. 1. The CS consists of a vacuum vessel (VV) and a copper shell (PS) mounted on its exterior surface. Specific ideal MHD instabilities, dubbed ideal kinks, can be stabilized by eddy currents generated within the CS. During plasma operation, eddy currents are induced in the CS and suppress Control system for MHD mode control Feedback control strategies play a crucial role in achieving better performance in the operation of the RFP devices [6], [7], [8]. The control schemes, such as raw mode control (RMC) and clean mode control (CMC) control strategies championed by RFX-mod, have been successfully implemented in various RFP systems. During the initial validation phase of the KTX real-time control system, the RMC control strategy was chosen for its simplicity. However, our ongoing research is focused on investigating Mode verification tests A series of mode verification tests have been carried out to evaluate the accuracy of the magnetic field generated by the feedback control system. The testing procedure involves the following steps: A predefined reference current signal is transmitted to the DPAs by a real-time control system operating in an open-loop configuration. The output current values from the DPAs and the strength of the radial magnetic field produced by the saddle coils at the boundary are measured and compared with Experiment During the initial phase of real-time feedback control experiments in KTX, the perturbed radial magnetic field at the device boundary is measured by saddle sensors, with data acquired by four MSACs, as depicted in the closed-loop feedback chain outlined in Fig. 11. These measured signals, b r θ i , ϕ j , are transmitted in real-time via optical fiber to the DSPC_1 module for processing and generation of 4 × 12 feedback current signals i f e e d θ i , ϕ j . These signals are then conveyed to the DSPC_2 module Conclusion The FPGA-based distributed real-time feedback system has been successfully designed, tested, and implemented for the first time on the KTX device. Harnessing the rapid computational capabilities of FPGA, the system satisfies the stringent real-time control requirements of the KTX device, the delivering response time is less than 50 μ s . Operating at a feedback control signal streaming rate of 25 kHz, this highly scalable system caters to the measurement needs of various EMPs. With successful CRediT authorship contribution statement Zhen Tao: Writing – review & editing, Conceptualization. Shuchen Song: Software, Hardware, Data curation. Hong Li: Writing – review & editing, Conceptualization. Adil Yolbarsop: Writing – review & editing, Conceptualization. Kezhu Song: Software, Hardware, Data curation. Jiahong Jiang: Software, Hardware, Data curation. Yuan Zhang: Project administration. Wentan Yan: Project administration. Zheng Chen: Project administration. Xianhao Rao: Project administration. Shunrong Ren: Project Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments The authors would like to express their sincere thanks to all members of the KTX group, as well as the State Key Laboratory of Particle Detection and Electronics at the University of Science and Technology of China and the Southwestern Institute of Physics, for their valuable contributions and support. This work was supported by the National Magnetic Confinement Fusion Energy Development Program of China [Grant No. 2017YFE0301702 ] and the National Natural Science Foundation of China [Grant No. References (27) Chen Y. The implementation of a data acquisition and service system based on HDF5 Fusion Eng. Des. (2016) Manduchi G. From distributed to multicore architecture in the RFX-mod real time control system Fusion Eng Des. (2014) Freidberg Jeffrey P. Ideal MHD (2014) Marrelli Lionello The reversed field pinch Nucl. Fusion (2021) Wang Z.R. et al. Physical understanding of the instability spectrum and the feedback control of resistive wall modes in reversed field pinch Nucl. Fusion (2011) Berkery J.W. Resistive wall mode instability at intermediate plasma rotation Phys. Rev. Lett. (2010) Frassinetti Lorenzo Resonant magnetic perturbation effect on tearing mode dynamics Nucl. Fusion (2010) Zanca Paolo Beyond the intelligent shell concept: The clean-mode-control Nucl. Fusion (2007) Brunsell Per R. Resistive wall mode feedback control in EXTRAP T2R with improved steady-state error and transient response Phys. Plasmas (2007) Tanaka Hiroyuki Effect on plasma performance of a single MHD mode feedback control in low-aspect-ratio RFP RELAX Plasma Fus. Res. (2014) Lorenzini Rita Self-organized helical equilibria as a new paradigm for ohmically heated fusion plasmas Nat. Phys. (2009) Martini S. Active MHD control at high currents in RFX-mod Nucl. Fusion (2007) Liu Wandong Progress of the Keda Torus eXperiment project in China: Design and mission Plasma Phys. Contr. F. (2014) View more references Cited by (0) Recommended articles (0) View full text © 2023 Elsevier B.V. All rights reserved. About ScienceDirect Remote access Shopping cart Advertise Contact and support Terms and conditions Privacy policy We use cookies to help provide and enhance our service and tailor content and ads. By continuing you agree to the use of cookies . Copyright © 2023 Elsevier B.V. or its licensors or contributors. ScienceDirect® is a registered trademark of Elsevier B.V. ScienceDirect® is a registered trademark of Elsevier B.V.
|
