Empty Banner

TRAINER - Tokamak plasmas daTa-dRiven identificAtIon and magNEtic contRol

WP3

Development of DRL-based tuning procedures to improve the robustness of model-based magnetic control algorithms

This WP aims at optimizing the control gains of the model-based plasma magnetic control algorithms currently proposed for ITER  by running nonlinear simulations with the fast version of CREATE-NL and by resorting to DDPG algorithms to improve the performance, by means of an optimization of the control gains. Indeed, control parameters are usually finely tuned using plasma linear models to optimize the performance around a given operating scenario. Optimization using a nonlinear model will contribute to improve the robustness of the classical model-based approaches.

The reference model-based control algorithms considered by this WP  are the one currently proposed for the ITER tokamak PCS. This WP will exploit the guidelines defined by WP1 to tune the DDPG hyper-parameters.


WP3 will include the single following task:

[T3.1] – Optimization of the controller gains for the plasma current, shape and VS system exploiting DDPG algorithms and fast nonlinear simulations with the nonlinear equilibrium code

 

The following deliverables are expected as outcomes of [WP3]:

[D7] – Report on the strategy to be adopted to tune control gains model-based plasma magnetic control algorithms by using DDPG for a given plasma scenario

 

Contacts

Street address:

Gianmaria De Tommasi

DIETI

Via Claudio, 21 - Naples - Italy - 80125

Email:

prin.trainer@unina.it