Identification of an Open-loop Plasma Vertical Position Using Fractional Order Dynamic Neural Network

TL;DR

This paper introduces a fractional order dynamic neural network to accurately identify the vertical position of plasma in a Tokamak, demonstrating improved performance over traditional integer order neural networks using real experimental data.

Contribution

It proposes a novel fractional order neural network model for plasma position identification, extending existing neural network approaches with fractional calculus.

Findings

FODNN outperforms integer order neural networks in plasma position prediction

The method effectively models complex plasma dynamics in Tokamaks

Numerical simulations align well with experimental data

Abstract

In order to identify complicated systems, more prominent and promising methods are needed among which we may refer to fractional order differential equations. The aim of this paper is to propose a fractional order nonlinear model to predict the vertical position of a plasma column system in a Tokamak by using real data from Damavand Tokamak. The system is identified based on a newly introduced fractional order dynamic neural network. The proposed fractional order dynamic neural network (FODNN) is an extension of the integer order dynamic neural network that employs the so called fractional-order operators. FODNN is implemented and comparison of the numerical simulation results with experimental results shows that performance of the proposed method by using fractional order neural network is preferred to the integer neural network.