INFORMATICS AND INFORMATION PROCESSING

Recovery of electron density signals beyond the operating range of the measuring instrument

N. V. Leshov^a, A. N. Shcherbak^b, M. G. Gorodnichev<sup>a

a</sup> Moscow Technical University of Communications and Informatics (MTUCI)
^b State Research Centre of the Russian Federation Troitsk Institute for Innovation and Fusion Research

Abstract: Machine learning models have been widely incorparated into control systems aimed at improving the operational efficiency of tokamaks. The training machine learning models requires substantial datasets. However, data collection is limited because experimental campaigns on tokamaks are prolonged in time. Furthermore, the amount of suitable training data may decrease due to the present of faulty diagnostic signals. Additionally, the frequency of faulty signal occurrences increases while initial operation of a new tokamak or specialized equipment. This work examines the possibility of recovering faulty signals using machine learning techniques. Particularly, we focus on recovering signals obtained beyond the operating range of measuring instruments. Thus, recovering such kind of signals should increase the volume of available training data, consequently enhancing the efficacy of machine learning-based model training.

Keywords: tokamak, plasma density, interferometry, artificial neural network, signal recovery.

UDC: 004.896

DOI: 10.33693/2313-223X-2025-12-3-152-159