Journal of Plasma Physics

Table of Contents - April 2012 - Volume 78, Issue 02  

Papers

Stochastic modeling of plasma mode forecasting in tokamak

SH. SAADATa1, M. SALEMa2, M. GHORANNEVISSa2 and P. KHORSHIDa3

a1 Faculty of Science, Ahvaz Branch, Islamic Azad University, Ahvaz, Iran (shsaadat@iauahvaz.ac.ir)

a2 Plasma Physics Research Center, Tehran Science & Research Branch, Islamic Azad University, P.O. Box 14665-678, Tehran, Iran

a3 Group Physics, Mashhad Branch, Islamic Azad University, Mashhad, Iran

Abstract

The structure of magnetohydrodynamic (MHD) modes has always been an interesting study in tokamaks. The mode number of tokamak plasma is the most important parameter, which plays a vital role in MHD instabilities. If it could be predicted, then the time of exerting external fields, such as feedback fields and Resonance Helical Field, could be obtained. Autoregressive Integrated Moving Average (ARIMA) and Seasonal Autoregressive Integrated Moving Average are useful models to predict stochastic processes. In this paper, we suggest using ARIMA model to forecast mode number. The ARIMA model shows correct mode number (m = 4) about 0.5 ms in IR-T1 tokamak and equations of Mirnov coil fluctuations are obtained. It is found that the recursive estimates of the ARIMA model parameters change as the plasma mode changes. A discriminator function has been proposed to determine plasma mode based on the recursive estimates of model parameters.

(Received June 16 2011)

(Revised September 08 2011)

(Accepted September 19 2011)

(Online publication November 11 2011)