29th Symposium on Fusion Technology (SOFT 2016)

O3A.3 Model identification of the temperature on the FTU liquid lithium limiter

6 Sep 2016, 17:20

20m

Congress Hall 2nd floor (Prague Congress Centre)

Board: 3

Oral D. Diagnostics, Data Acquisition and Remote Participation O3A

Speaker

Claudia Corradino (DIEEI)

Description

One of the main research lines currently investigated within the FTU programs is the possibility to adopt a technology based on liquid metals as first plasma wall. More particularly, the main attention has been devoted to the analysis of plasma performances when using a liquid lithium limiter (LLL) device. The control of the limiter surface temperature reveals to be a fundamental aspect of the LLL operativity, hence a model able to link the different physical observables to the thermal behavior of the LLL has to be identified. The temperature over the surface of the LLL in FTU is monitored through an infrared thermocamera located inside the vacuum chamber. We performed a complete analysis of the datasets collected at FTU during each experiment, evaluating the correlation coefficients with respect to the spatio temporal distribution of the temperature over the limiter surface as measured by the infrared camera. The proposed approach relies on the training of an artificial neural network (ANN) with the aim of interpolating the dynamical relationship between the identified quantities and the thermal process with a nonlinear, autoregressive moving average model. The evaluation of these preliminary results, which are able to show an overall good estimation of the dynamical model regulating the thermal process over the limiter surface, suggests that the highly nonlinear dynamics regulating the interplay be tween plasma and thermal behavior of the limiter can be successfully modeled with an ANN.