Inspection of Disruptive Behaviours at JET using Generative Topographic Maps

Tokamaks are the most promising configuration of magnetic confinement fusion devices. However, a physical phenomenon that leads the plasma out to its operational bounds, called disruption, remains unavoidable. Disruptions cause the abrupt termination of the discharge and in addition to affecting the execution of the research program, they can constitute a risk for the structural integrity of the machine. Their occurrences have proven to be unavoidable, particularly in high performance scenarios. In this article two important aspects that can facilitate the better understanding of the phenomenon are presented. First, the selection of the physical parameters and their main characteristics related to disruptions are reviewed. This feature extraction procedure consists in the selection of the most adequate plasma measurements and the processing of each selected signal to extract and condense the disruptive-related characteristics. Second, the application of Generative To-pographic Mapping (GTM) to visualize and compare disruptive and non disruptive experiments at different times is shown. The resulting maps are aimed to evidence the evolution of the phenomenon, since it is un-recognizable till it can be distinguished. The identification of the instant when precursors of disruptions can be noticed is highly relevant in nuclear fusion since it determines the time margin the control systems have to apply mitigation or avoidance actions.