An optimal feedback approach to particle source identification in tokamaks

A new identification technique is proposed to study the plasma phenomena taking place in the so-called scrape-off layer and their correlation with the lower hybrid radio frequency antenna in the experimental nuclear fusion tokamak reactor Tore Supra. A deeper knowledge of the plasma behavior in this region would contribute to the achievement of steady-state controlled thermonuclear fusion for power generation. The proposed approach relies on the design of a feedback/feedforward optimized architecture to solve a blind identification problem. While our paper is mostly focused on experimental results for the studied application, it can provide valuable insights on input estimation and model validation for transport phenomena described by partial differential equations.