• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Plasma physics and controlled fusion >Development of real-time diagnostics and feedback algorithms for JET in view of the next step
【24h】

Development of real-time diagnostics and feedback algorithms for JET in view of the next step

机译：下一步，为JET开发实时诊断和反馈算法

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Real-time control of many plasma parameters will be an essential aspect in the development of reliable high performance operation of next step tokamaks. The main prerequisites for any feedback scheme are the precise real-time determination of the quantities to be controlled, requiring top quality and highly reliable diagnostics, and the availability of robust control algorithms.A new set of real-time diagnostics was recently implemented on JET to prove the feasibility of determining, with high accuracy and time resolution, the most important plasma quantities. Some of the signals now routinely provided in real time at JET are: (i) the internal inductance and the main confinement quantities obtained by calculating the Shafranov integrals from the pick-up coils with 2 ms time resolution; (ii) the electron temperature profile, from electron cylotron emission every 10 ms; (iii) the ion temperature and plasma toroidal velocity profiles, from charge exchange recombination spectroscopy, provided every 50 ms; and (iv) the safety factor profile, derived from the inversion of the polarimetric line integrals every 2 ms. With regard to feedback algorithms, new model-based controllers were developed to allow a more robust control of several plasma parameters.With these new tools, several real-time schemes were implemented, among which the most significant is the simultaneous control of the safety factor and the plasma pressure profiles using the additional heating systems (LH, NBI, ICRH) as actuators. The control strategy adopted in this case consists of a multi-variable model-based technique, which was implemented as a truncated singular value decomposition of an integral operator. This approach is considered essential for systems like tokamak machines, characterized by a strong mutual dependence of the various parameters and the distributed nature of the quantities, the plasma profiles, to be controlled. First encouraging results were also obtained using non-algorithmic methods like neural networks, which have been successfully applied to non-linear and ill-posed problems, for example the determination of the divertor radiated power.The real-time hardware and software architectures adopted are also described with particular attention to their relevance to ITER.
机译：在下一步托卡马克可靠的高性能操作开发中，许多血浆参数的实时控制将是必不可少的方面。任何反馈方案的主要先决条件是精确实时确定要控制的数量，需要最高质量和高度可靠的诊断以及鲁棒的控制算法的可用性。最近在JET上实施了一套新的实时诊断证明以高精度和高时间分辨率确定最重要的血浆量的可行性。现在在JET上实时实时提供的一些信号是：（i）内部电感和主要约束量，该内部电感和主要约束量是通过以2 ms的时间分辨率从拾取线圈计算Shafranov积分而获得的; （ii）电子温度曲线，每10毫秒由电子回正电子发射一次; （iii）每50毫秒提供一次来自电荷交换复合光谱的离子温度和等离子体环面速度分布图; （iv）安全系数曲线，由每2 ms极化线积分的反转得出。关于反馈算法，开发了基于模型的新控制器以允许更稳健地控制几个血浆参数。使用这些新工具，实现了多个实时方案，其中最重要的是对安全系数的同时控制以及使用附加加热系统（LH，NBI，ICRH）作为执行器的等离子压力曲线。在这种情况下采用的控制策略由基于变量的基于模型的技术组成，该技术被实现为积分算子的截断奇异值分解。对于像托卡马克机这样的系统，该方法被认为是必不可少的，其特点是各种参数之间的强烈相互依赖性以及要控制的量的分布特性（血浆分布）。使用神经网络等非算法方法也获得了令人鼓舞的结果，这些方法已成功应用于非线性和不适定问题，例如确定偏滤器的辐射功率，并采用了实时硬件和软件架构还特别注意了它们与ITER的相关性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号