Artificial Neural Networks and Multi-Class Support Vector Machines for Classifying Magnetic Measurements in Tokamak Reactors

This paper is mainly concerned with the application of a novel technique of data interpretation for classifying measurements of plasma columns in Tokamak reactors for nuclear fusion applications. The proposed method exploits several concepts derived from soft computing theory. In particular, Artificial Neural Networks and Multi-Class Support Vector Machines have been exploited to classify magnetic variables useful to determine shape and position of the plasma with a reduced computational complexity. The proposed technique is used to analyze simulated databases of plasma equilibria based on ITER geometry configuration. As well as demonstrating the successful recovery of scalar equilibrium parameters, we show that the technique can yield practical advantages compared with earlier methods.

Authors:

• A. Greco
• N. Mammone
• F.C. Morabito
• M.Versaci

Keywords:

• Tokamak
• Classification
• Artificial Neural Network
• Support Vector Machines.

References:

[1] J. Wesson, "Tokamaks", Oxford Science Pub., 1987.
[2] M. Matsukawa, et al., "Application of Regression Analysis to Deriving
Measurements Formulas for Feedback Control of Plasma Shape in JT-
60", Plasma Physics and Controlled Fusion, vol. 34, N┬░ 6, pp. 907-921,
1992.
[3] F.C. Morabito, M. Versaci, "Fuzzy Inference Systems (FIS) for
Classification Identification of Plasma Columns in Tokamak Reactors",
International Journal of Chaos Theory and Applications, Vol. 5, N┬░1, pp.
11-40, 2000.
[4] P.J. McCarthy, F.C. Morabito, "Function Parametrization and Artificial
Neural Networks for Equilibrium Parameters Recovery in ASDEXUpgrade:
a Comparison", Journal of Applied Electromagnetics and
Mechanics, Vol. 9, pp. 1-31, 1997.
[5] F.C. Morabito, M. Campolo, "Are Hybrid Fuzzy-Neural Systems
Actually Useful in Plasma Engineering?, Proc. Of the 9th Italian
Workshop on Neural Nets, M. Marinaro and R. Tagliaferri Eds.,
WIRN-97, Vietri S/M, pp. 22-24, 1997.
[6] F.C. Morabito, M. Versaci, "A Fuzzy-Neural Approach to Real Time
Plasma Boundary Reconstruction in Tokamak Reactors", IEEE ICNN
International Conference on Neural Networks, Houston, Texas, pp. 43-
47, June 1997.
[7] F.C. Morabito et al. "Final report on EFDA Study Contract FU05 CT
2002-00162 (EFDA 02-1001)", 2002.
[8] C. M. Bishop, "Neural Network for Pattern recognition", Clarendon
Press, Oxford, 1995.
[9] F.C. Morabito, et al., "On Line Plasma Shape Identification in a
Tokamak Reactor Via Neural Network", Proc. of V Italian Workshop on
Neural Networks, Word Scientific Publishing, p. 349, 1992.
[10] B.Cortes, V.N. Vapnik, Support vector networks, Machine Learning
Vol. 20, pp. 273-297, 1995.
[11] C. W. Hsu, C.-J. Lin, "A comparison of methods for Multi-class Support
Vector Machines", Department of Computer Science and Information
Engineering, National Taiwan University.