Applications of a novel model-based real-time observer for electron density profile control experiments in TCV Journal: Nuclear Fusion
F. Pastore, O. Sauter, F. Felici, D. Kropackova, A. Balestri, C. Galperti, O. Kudlacek, K. Lee, A. Pau, T. Ravensbergen, S. Van Mulders, B. Vincent, N.M.T. Vu, the TCV team, and the EUROfusion Tokamak Exploitation Team · 2026
Read Applications of a novel model-based real-time observer for electron density profile control experiments in TCV Journal: Nuclear Fusion by F. Pastore, O. Sauter, F. Felici, D. Kropackova, A. Balestri, C. Galperti, O. Kudlacek, K. Lee, A. Pau, T. Ravensbergen, S. Van Mulders, B. Vincent, N.M.T. Vu, the TCV team, and the EUROfusion Tokamak Exploitation Team (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-26310.txt.
Introduction
Pastore and co-authors test a real-time observer that tracks electron density profiles in the TCV tokamak. Density control matters for confinement, heating, exhaust, and staying under the density limit. The RAPDENS observer now feeds the TCV control system in detachment, L-mode, and H-mode runs. It holds line-averaged density inside the last closed flux surface while rejecting stray divertor signals. It also supports local control of central density and live estimates of transport traits.
Catalog record
| CREATOR | F. Pastore, O. Sauter, F. Felici, D. Kropackova, A. Balestri, C. Galperti, O. Kudlacek, K. Lee, A. Pau, T. Ravensbergen, S. Van Mulders, B. Vincent, N.M.T. Vu, the TCV team, and the EUROfusion Tokamak Exploitation Team |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 106714 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.26310
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Applications of a novel model-based real-time observer for electron density profile control experiments in TCV Journal: Nuclear Fusion F. Pastore, O. Sauter, F. Felici, D. Kropackova, A. Balestri, C. Galperti, O. Kudlacek, K. Lee, A. Pau, T. Ravensbergen, S. Van Mulders, B. Vincent, N.M.T. Vu, the TCV team, and the EUROfusion Tokamak Exploitation Team Abstract The main aspects of real-time control of tokamak plasmas today are ensuring a stationary state around a highly performing target, proximity control to avoid disruptions, and controlling the access (ramp-up) and termination (ramp-down) of the shot. Within these applications, the real-time estimation and control of the electron density is fundamental for monitoring plasma confinement, heating efficiency, and control exhaust conditions, impurity concentration, fusion power production, and proximity to the so-called density limit 21 . Building on the integration of a novel multi-rate observer 50 based on RAPDENS 6 into the TCV plasma control system, this follow-up study explores its application to density profile control for detachment studies, ECH- and NBH- L-mode plasmas, and high-performance H-mode scenarios. Experiments on TCV demonstrate the observer’s capability to support detachment studies in complex divertor geometries by accurately controlling the line-averaged density within the last-closed flux surface while avoiding the pick-up of the far-infrared interferometer signal attributed to the Scrape-Off Layer electron density in the divertor region. The estimated electron density profile from the observer enables local control of the central density in an L-mode plasma with a mix of ECH and NBH, to operate below cutoff conditions. The effect of these heating schemes on the electron density profile peaking is analyzed and treated as a disturbance to the control task. Real-time estimation and adoption of the unknown, time-varying particle transport coefficient profiles, such as the ratio between the electron pinch velocity and diffusion coefficient, is carried out to further improve the predictive capabilities and spatial accuracy of the control-oriented model. The turbulent transport observed in the experimental scenario is characterized using linear and non-linear gyrokinetic simulations with the GENE code to verify the observed particle pumpout in these experiments. Relying on the good spatial resolution of the density profile reconstruction at the top of the pedestal, the simultaneous control of the edge-normalized density and toroidal beta of high-performance H-mode plasmas is demonstrated. This approach yields good confinement properties, scenario reproducibility, and an edge-density metric independent of the diagnostics available on the tokamak, while avoiding density limits, plasma disruptions, and the propagation of diagnostic faults in the control scheme. 1 Introduction & Motivation Estimation and control of plasma kinetic quantities, such as plasma electron density, temperature, and current density profile, have gained increased relevance over the years. To maximize plasma fusion output, avoid operational limits and MHD instabilities, and keep the plasma scenario controlled over stable trajectories, a precise real-time assessment of these quantities is deemed necessary 63 , 4 . Future fusion power plants represent a challenging environment for many physics, technological, and engineering problems. The nominal operation point of the power plant is bound to lie close to physical and technological limits for an economical scalability 71 and electric cost competitiveness. The operation of the fusion reactor is also constrained by a limited set of diagnostics, due to the harsh reactor conditions expected on a burning plasma device. The Plasma Control System (PCS) of the reactor has to be designed to guarantee robust and reliable operation, integrate and orchestrate the different signals coming from diagnostics, actuators, and algorithms, handle off-normal events, and ensure robust performance in real time in case of diagnostics failures or unavailability. The integration of plasma state observers in the PCS, designed to predict and reconstruct the plasma state, can be exploited to provide robust performance by combining different diagnostics and models in real time in an integrated fashion. The reconstructed state can then inform the supervisor and off-normal event manager (e.g. SAMONE 69 ) and track different plasma quantities to ensure that the operational limits are not exceeded and lie in a high-performance operational space. 1.1 State of the art - kinetic profiles estimators for offline and real-time applications There are multiple examples of observers developed in recent years for offline and real-time plasma kinetic state estimation. These observers perform the reconstruction of plasma kinetic profiles, exploiting the available information provided by diagnostics, equilibrium codes, and actuator signals. Model-based approaches, relying on the solution of flux-surface averaged equations for the evolution of kinetic profiles, have been developed and tested on different plasma devices over the last decade 14 , 43 , 11 , 49 , 44 , 50 , 31 , 66 , 67 . Data-driven approaches, based on machine learning techniques, have been recently designed and deployed on fusion machines for the real-time reconstruction of kinetic quantities 61 , 23 , 28 and the enhancement of diagnostic data 22 . A model-based approach to predict and estimate the electron density profile has been introduced in 6 . RAPDENS, the RApid Plasma DENsity Simulator, has been formulated, programmed, and tested in real time on TCV and ASDEX Upgrade. It incorporates a control-oriented model to describe the dynamics of the flux-surface averaged electron density profile in a tokamak. The modeling of the neutrals in the vacuum chamber and reactor wall has been introduced with two dedicated 0-D inventories to describe the ionization, recycling, recombination, flux of particles to the Scrape-Off Layer (SOL) and plasma fueling through different actuators, such as gas puffing, Neutral Beam Injection (NBI), and pellet fueling. The real-time-capable state observer is based on the Extended Kalman Filter (EKF) 27 technique, where diagnostic measurements related to the electron density or neutrals are incorporated into the prediction of the plasma state provided by the non-linear physics model. A first real-time application of the code on TCV and ASDEX Upgrade has been shown in 5 , where integral measurements of interfermeters and Bremsstrahlung are integrated in the EKF. RAPDENS follows a similar approach as the RAPTOR code 15 , the RApid Plasma Transport simulatOR, developed to predict and estimate the plasma density, temperature, and current density profiles. Recent results on the estimation of the electron temperature T_{e} , density n_{e} , and current density profiles for ITER 66 using a synthetic diagnostic model for Thomson scattering (TS) system have been proposed, where the simultaneous estimation of the plasma state and unknown plasma parameters has been carried out within the formulation of RAPTOR EKF. The improved state estimator is then leveraged for offline Kinetic Equilibrium Reconstruction of TCV plasmas 67 , where a self-consistent solution of the magnetic equilibrium problem and internal kinetic state is achieved, leading to accurate reconstruction of the safety factor profile and dynamical evolution of the sawtooth instability. In 50 , a novel multi-rate EKF based on the RAPDENS model has been presented, fusing high-frequency far-infrared interferometer (FIR) data and low-frequency TS data to provide an estimate of the electron density profile with high spatial and temporal accuracy. The model-based observer, integrated in the TCV PCS 18 , has been further refined and exploited to assist various experimental missions that can benefit from direct control of the electron density profile. These experiments are used to validate the proposed framework for high-performance, high-density plasmas, with various heating sources, as well as demonstrating how