Design, Simulation, and Analysis of Reactor Relevant Plasma Control State Estimator

Effective plasma control is essential to fusion reactor operation; however, reduced diagnostic availability and performance relative to current experiments are expected to limit control capabilities. A reactor-relevant plasma state estimator based on the Ensemble Kalman Filter (EnKF) has been designed and demonstrated in simulation to address these challenges. The estimator assimilates diagnostic data via the EnKF to provide real-time control parameters along with uncertainty quantification. Unlike traditional approaches, it integrates diagnostic inputs directly, bypassing separate diagnostic interpretation. The state estimator is physics-informed, employing simplified models for equilibrium and profile evolution, while remaining extensible to more sophisticated models as required. The state estimator has been tested and demonstrated in TokSys/GSevolve simulations using a hypothetical ITER variant equipped only with reflectometers and ECE diagnostics, and with inductive sensing limited to total plasma current (Ip). High-accuracy estimation of key control parameters—such as vertical position and gap distances—has been achieved for the ITER baseline scenario, spanning ramp-up to flattop phases. The estimator is further applied in closed-loop control within GSevolve simulations for the same scenario. The robust performance of the state estimator not only supports future fusion pilot plant concepts, but also enables control-informed design, where estimators can assess the minimal viable diagnostic set required for effective control—an important step toward demonstrating economic feasibility.