Neural-Network accelerated TGLF model for ITER scenario modeling

Preparation for ITER operation relies on our ability to efficiently predict the plasma confinement with high physics fidelity. High-fidelity turbulent transport models such as TGLF remain one of the major bottlenecks in this process. To accelerate prediction of the turbulent transport, the neural network (NN) approach of TGLF (TGLF-NN) described in [Meneghini NF 2017] has been extended for ITER. A large database of ITER TGLF simulations was assembled for both the commissioning phase (H only) and the nuclear phase (D+T and He ash). The original NN implementation has been re-written to leverage state-of-the-art machine learning libraries (e.g. Tensorflow). Tuning of the NN model hyper-parameters (topology and training parameters) was carried out on GPU enabled clusters, both using Gaussian process based optimization, and random sampling of the configuration space. Dimensional reduction with auto-encoders, and training on a latent-space data-set was also investigated. Results of coupled core-pedestal ITER simulations leveraging the latest EPED1-NN and TGLF-NN models will be presented.