Three machine learning techniques (multilayer perceptron, random forest, and Gaussian process) provide fast surrogate models for lower hybrid current drive (LHCD) simulations. A single GENRAY/CQL3D simulation without radial diffusion of fast electrons requires several minutes of wall-clock time to complete, which is acceptable for many purposes, but too slow for integrated modeling and real-time control applications. More accurate simulations with fast electron diffusion are even slower, requiring multiple hours of run time with parallel processing. The machine learning models use a database of 16,000+ GEN-RAY/CQL3D simulations for training, validation, and testing. Latin hypercube sampling methods implemented in πScope ensure that the database covers the range of 9 input parameters (ne0, Te0, Ip, Bt, R0, n∥︀, Ze f f, Vloop, PLHCD) with sufficient density in all regions of parameter space. The surrogate models reduce the computation time from minutes-hours to ms with high accuracy across the input parameter space. Data-driven surrogate models also allow for solving inverse and "lateral"problems. A surrogate model for the inverse problem maps from a desired current drive or power deposition profile to a set of input parameters that would result in such a profile, while a surrogate model for the lateral problem maps from a measured experimental quantity such as hard x-ray emission to a current drive or power deposition profile. The πScope database creation workflow is flexible and applicable to other RF simulation codes such as TORIC.

Three machine learning techniques (multilayer perceptron, random forest and Gaussian process) provide fast surrogate models for lower hybrid current drive (LHCD) simulations. A single GENRAY/CQL3D simulation without radial diffusion of fast electrons requires several minutes of wall-clock time to complete, which is acceptable for many purposes, but too slow for integrated modelling and real-Time control applications. The machine learning models use a database of more than 16Â 000 GENRAY/CQL3D simulations for training, validation and testing. Latin hypercube sampling methods ensure that the database covers the range of nine input parameters (, and) with sufficient density in all regions of parameter space. The surrogate models reduce the inference time from minutes to ms with high accuracy across the input parameter space.

A surrogate model of the wave absorption in the ion cyclotron range of frequencies is presented. The model is trained to capture the physics of 1D electron and ion power absorption profiles for both the high harmonic fast wave scheme in NSTX, and the minority heating scheme in WEST. The surrogate models, based on both the random forest regressor and the multilayer perceptron algorithms, reduce inference time of 1D power absorption profiles from 1-5 minutes required by TORIC to ∼50 µs with high accuracy (i.e. R2 = 0.71−0.96).

A real-time capable core Ion Cyclotron Range of Frequencies (ICRF) heating model on NSTX and WEST is developed. The model is based on two nonlinear regression algorithms, the random forest ensemble of decision trees and the multilayer perceptron neural network. The algorithms are trained on TORIC ICRF spectrum solver simulations of the expected flat-top operation scenarios in NSTX and WEST assuming Maxwellian plasmas. The surrogate models are shown to successfully capture the multi-species core ICRF power absorption predicted by the original model for the high harmonic fast wave and the ion cyclotron minority heating schemes while reducing the computational time by six orders of magnitude. Although these models can be expanded, the achieved regression scoring, computational efficiency and increased model robustness suggest these strategies can be implemented into integrated modeling frameworks for real-time control applications.

Application of lower hybrid (LH) current drive in tokamak plasmas can induce both co- and countercurrent directed changes in toroidal rotation, depending on the core q profile. For discharges with q(0) <1, rotation increments in the countercurrent direction are observed. If the LH-driven current is sufficient to suppress sawteeth and increase q(0) above unity, the core toroidal rotation change is in the cocurrent direction. This change in sign of the rotation increment is consistent with a change in sign of the residual stress (the divergence of which constitutes an intrinsic torque that drives the flow) through its dependence on magnetic shear.

Research on the spherical torus (or spherical tokamak) (ST) is being pursued to explore the scientific benefits of modifying the field line structue fro that in more moderate aspect ratio devices. The ST experiments are being conducted in various US research facilities. The area of power and particle handling is expected to be challenging because of the higher power density expected in the ST relative to that in conventional aspect-ratio tokamaks.

A major research goal of the national spherical torus experiment is establishing long-pulse, high beta, high confinement operation and its physics basis. This research has been enabled by facility capabilities developed during 2001 and 2002, including neutral beam (up to 7 MW) and high harmonic fast wave (HHFW) heating (up to 6 MW), toroidal fields up to 6 kG, plasma currents up to 1.5 MA, flexible shape control, and wall preparation techniques. These capabilities have enabled the generation of plasmas with βT ≡

/(BT02/2μ0) of up to 35%. Normalized beta values often exceed the no-wall limit, and studies suggest that passive wall mode stabilization enables this for H mode plasmas with broad pressure profiles. The viability of long, high bootstrap current fraction operations has been established for ELMing H mode plasmas with toroidal beta values in excess of 15% and sustained for several current relaxation times. Improvements in wall conditioning and fuelling are likely contributing to a reduction in H mode power thresholds. Electron thermal conduction is the dominant thermal loss channel is auxiliary heated plasmas examined thus far. HHFW effectively heats electrons, and its acceleration of fast beam ions has been observed. Evidence for HHFW current drive is obtained by comparison of the loop voltage evolution in plasmas with matched density and temperature profiles but varying phases of launched HHFW waves. Studies of emissions from electron Bernstein waves indicate a density scale length dependence of their transmission across the upper hybrid resonance near the plasma edge that is consistent with theoretical predictions. A peak heat flux to the divertor targets of 10 MW m-2 has been measured in the H mode, with large asymmetries being observed in the power deposition between the inner and outer strike points. Non-inductive plasma startup studies have focused on coaxial helicity injection. With this technique, toroidal currents up to 400 kA have been driven, and studies to assess flux closure and coupling to other current drive techniques have begun.