Since 18 of December 2019 conferences.iaea.org uses Nucleus credentials. Visit our help pages for information on how to Register and Sign-in using Nucleus.
22-27 October 2018
Mahatma Mandir Conference Centre
Asia/Kolkata timezone
CONFERENCE MATERIAL NOW AVAILABLE!

Physics-model-based Real-time Optimization for the Development of Steady-state Scenarios at DIII-D

25 Oct 2018, 14:00
4h 45m
Mahatma Mandir Conference Centre

Mahatma Mandir Conference Centre

Gandhinagar (nearest Airport: Ahmedabad), India
Poster P6 Posters

Speaker

Prof. Eugenio Schuster (Lehigh University)

Description

Recent and ongoing experiments on DIII-D demonstrate the potential of model-based real-time optimization for the realization of advanced steady-state scenarios by tightly regulating the $q$ profile and $\beta_N$ (or the plasma energy $W$) simultaneously. A primary goal for the DIII-D research program over the next five years is to develop the physics basis for a high $q$ ($q_{\text{min}}>2$), high $\beta_N$, steady-state scenario (fully relaxed plasma state where the current is entirely noninductive) that can serve as the basis for future steady-state burning plasmas. Various approaches are being considered to maximize both the bootstrap current and the noninductive current-drive contributions, so that fully noninductive ($f_{NI} = 1$) discharges can be obtained for several resistive current diffusion times. It is anticipated that the upcoming upgrades to DIII-D, including an additional off-axis neutral beam injection (NBI) system, will provide sufficient auxiliary current drive to maintain fully noninductive plasmas at high $\beta_N$. However, much work is still necessary to investigate MHD stability, adequate confinement, and early achievement and sustainment of the steady-state condition. The capability of combined $q$-profile and $\beta_N$ control to enable access to and repeatability of steady-state scenarios for $q_{\text{min}}>1.4$ discharges has been assessed in DIII-D experiments. To steer the plasma to the desired state, a model predictive control approach to both $q$-profile and $\beta_N$ regulation numerically solves successive optimization problems in real time over a receding time horizon by exploiting efficient quadratic programming techniques. A key advantage of this control approach is that it allows for explicit incorporation of plasma-state/actuator constraints to prevent the controller from driving the plasma outside of stability/performance limits and obtain, as closely as possible, steady state conditions. Experimental results demonstrate the effectiveness of the real-time optimization scheme to consistently achieve the desired scenarios at predefined times and suggest that control-oriented model-based scenario planning in combination with real-time optimization can play a crucial role in exploring stability limits of advanced steady-state scenarios.
Country or International Organization United States of America
Paper Number EX/P6-39

Primary author

Prof. Eugenio Schuster (Lehigh University)

Co-authors

Mr Andres Pajares (Lehigh University) Mr Benjamin Penaflor (General Atomics) Dr Christopher T. Holcomb (Lawrence Livermore National Laboratory) Dr David Humphreys (General Atomics) Dr John Ferron (General Atomics) Dr Michael Walker (General Atomics) Mr Robert Johnson (General Atomics) Dr William Wehner (Lehigh University)

Presentation Materials

Paper