Two-layer perceptron for fast plasma density evaluation in broadband reflectometry

Broadband reflectometry is a suitable diagnostic to measure plasma density profiles in nuclear fusion devices (e.g. tokamaks). To determine each value of the profile an Abel inversion integral has to be numerically solved. Unfortunately this is a computationally demanding operation that prevents the use of the diagnostic in real-time applications. In this paper we present a new approach to the evaluation of density profiles based on a two-layer perceptron that is faster than the Abel numerical integration. In addition, it has sufficient accuracy for most applications utilized in the tokamak's real-time control.