Community Research and Development Information Service - CORDIS

Abstract

Four different techniques for the fast analysis of multichannel data in plasma physics are discussed. All four of these techniques are general and sufficiently fast to be used in real time applications. Function parametrisation, canonical correlation analysis and a neural network of the multi-layer perceptron (MLP) type are compared to a unique linear mapping based on the singular value decomposition, which is used as a reference. Applications deal with the identification of the plasma boundary and global plasma parameters in the DIII-D and Tore Supra by using magnetic measurements. The results of a MLP-1 neural network, employed for the real time plasma position determination in Tore Supra, are presented.

Additional information

Authors: WIJNANDS T, CEA, Département de Recherches sur la Fusion Contrôlée, CEN Cadarache, Saint-Paul-lez-Durance (FR);PARLANOE F, CEA, Département de Recherches sur la Fusion Contrôlée, CEN Cadarache, Saint-Paul-lez-Durance (FR);COUTURIER B, CEA, Département de Recherches sur la Fusion Contrôlée, CEN Cadarache, Saint-Paul-lez-Durance (FR);MOULIN D, CEA, Département de Recherches sur la Fusion Contrôlée, CEN Cadarache, Saint-Paul-lez-Durance (FR)
Bibliographic Reference: Report: EUR-CEA-FC-1542 EN (1995) 38 pp.
Availability: Available from CEA, Département de Recherches sur la Fusion Contrôlée, Saint-Paul-lez-Durance (FR)
Record Number: 199510531 / Last updated on: 1995-04-11
Category: PUBLICATION
Original language: en
Available languages: en