pith. sign in

arxiv: 2604.26476 · v1 · submitted 2026-04-29 · 📡 eess.SY · cs.SY

Fuelling fusion plasmas with pellets: Can neuromorphic control outperform Sigma-Delta modulation?

L.L.T.C. Jansen , E. Petri , M. van Berkel , W.P.M.H. Heemels This is my paper

Pith reviewed 2026-05-07 11:24 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords pellet injectionfusion plasma controlneuromorphic controlsigma-delta modulationhybrid systemsstability analysistokamak fuellingdensity regulation
0
0 comments X

The pith

A neuromorphic controller treats pellet injections as neural spikes to regulate fusion plasma density and yields explicit parameter constraints for practical stability.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper builds a hybrid dynamical model for pellet fuelling in tokamaks, where discrete frozen-fuel injections create sudden density jumps inside continuous plasma evolution. It introduces a neuromorphic controller that fires pellets when an internal state crosses a threshold, mirroring integrate-and-fire neurons. For comparison it also models the sigma-delta modulator currently used in experiments. Explicit inequalities on pellet size, firing rate limits and controller gains are derived that guarantee the density stays inside a prescribed band around the target. Simulations confirm that both controllers meet the stability bounds when the inequalities hold and illustrate how the neuromorphic version exploits the hybrid character of the problem.

Core claim

We propose a formal hybrid model for the pellet fuelling process and a neuromorphic controller that treats pellets much like spikes as in a brain-like system. We also develop a hybrid model of sigma-delta modulation and derive explicit actuator and controller parameter constraints that lead to practical stability guarantees for both. Numerical simulations compare the controller variants and validate the theoretical results.

What carries the argument

The hybrid system model that interleaves continuous plasma density dynamics with instantaneous jumps at each pellet injection, together with the neuromorphic controller that generates discrete firing events from a state-dependent threshold.

If this is right

  • When actuator and controller parameters satisfy the stated inequalities, density remains inside a prescribed neighbourhood of the reference value for both neuromorphic and sigma-delta schemes.
  • The neuromorphic controller directly exploits the discrete-continuous structure of pellet injection without requiring continuous actuation.
  • The same stability conditions supply concrete tuning rules for choosing pellet size and firing thresholds in tokamak operation.
  • Simulation results show that the neuromorphic variant can achieve regulation performance at least comparable to sigma-delta while remaining computationally lightweight.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the derived constraints remain valid under measurement noise and pellet delivery delays present in real tokamaks, the neuromorphic approach could reduce the need for high-bandwidth feedback loops.
  • The same hybrid modelling and stability technique could be applied to other discrete fuelling or heating actuators in fusion devices.
  • Varying the internal state dynamics of the neuromorphic controller might allow explicit optimisation for minimal pellet consumption while preserving the stability bounds.

Load-bearing premise

The hybrid model accurately captures the instantaneous density jump caused by each pellet and the subsequent continuous plasma evolution on the relevant time scales.

What would settle it

A controlled experiment or high-fidelity simulation in which plasma density leaves the predicted bounded region even though all derived actuator and controller inequalities are satisfied would falsify the stability claims.

Figures

Figures reproduced from arXiv: 2604.26476 by E. Petri, L.L.T.C. Jansen, M. van Berkel, W.P.M.H. Heemels.

Figure 1
Figure 1. Figure 1: System with spiking controller (NM or SDM), refer view at source ↗
Figure 2
Figure 2. Figure 2: Numerical simulation of the neurmorphic controller view at source ↗
Figure 4
Figure 4. Figure 4: System with saturation block to limit the input to the view at source ↗
Figure 6
Figure 6. Figure 6: Numerical simulation of SDM with input clipping view at source ↗
Figure 7
Figure 7. Figure 7: Numerical simulation of SDM with adjusted jump view at source ↗
read the original abstract

Nuclear fusion is a promising clean energy source in which deuterium and tritium fuse inside a magnetically confined plasma in a tokamak, releasing energy. A key challenge on the route to practical nuclear fusion is the control of the plasma density which has to be done through adding fuel in the form of deuterium and tritium to the plasma. Pellet injection, firing frozen fuel into the plasma, is used to accomplish this. Since the injection of a pellet causes an almost instantaneous increase in particle density compared to the time scales of the plasma dynamics, the problem is of a hybrid nature in which continuous plasma dynamics are interrupted by discrete bursts of particles. In this paper, we propose a formal hybrid model for this fuelling process and we propose a new, neuron-inspired control method that treats pellets much like spikes as in a brain-like system. The neuromorphic controller offers a lightweight solution that naturally fits the hybrid character of pellet fuelling. For comparison, we also develop a hybrid model of sigma-delta modulation, which is used in current tokamaks. For both the neuromorphic controller and the sigma-delta modulation we present formal analysis results for this control problem in nuclear fusion. We derive explicit actuator and controller parameter constraints, key for controller tuning, that lead to practical stability guarantees. Numerical simulations compare the different controller variants and validate the theoretical results.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper develops a hybrid dynamical system model for pellet fuelling in tokamak plasmas, in which continuous plasma density evolution is interrupted by discrete near-instantaneous density jumps from pellet injections. It proposes a neuromorphic spiking controller for pellet actuation, contrasts it with a hybrid sigma-delta modulator, derives explicit actuator and controller parameter constraints that yield practical stability guarantees for both, and supports the claims via formal analysis and numerical simulations.

Significance. If the instantaneous-jump hybrid model is sufficiently faithful on the relevant timescales, the explicit tuning constraints and formal stability results could supply a lightweight, hybrid-native control method for density regulation in fusion devices, with the neuromorphic approach offering potential computational advantages. The provision of formal analysis yielding concrete parameter bounds and the comparative simulations are clear strengths.

major comments (2)
  1. [§2] §2 (hybrid model): the assumption that each pellet produces an instantaneous density jump interrupting continuous plasma dynamics is load-bearing for all subsequent stability analysis and the derived parameter constraints. The manuscript supplies only closed-loop simulations inside this model and provides neither experimental tokamak pellet data nor error bounds quantifying the effect of finite ablation/deposition profiles on the stability margins.
  2. [§5] §5 (stability analysis): the practical stability guarantees and explicit actuator/controller constraints are obtained under the hybrid model; without sensitivity analysis to model mismatch (e.g., non-instantaneous jumps), it remains unclear whether the tuning rules remain valid when the continuous inter-pellet evolution deviates from the assumed dynamics.
minor comments (2)
  1. [§2] Notation for the hybrid state and jump map should be introduced with a single consistent definition early in the model section to avoid later ambiguity.
  2. [§6] Simulation figures would be clearer if they explicitly labeled the neuromorphic versus sigma-delta traces and included a table of the exact parameter values satisfying the derived constraints.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the thoughtful and constructive report. The comments highlight important limitations of the modeling assumptions, which we address point by point below while preserving the scope of the theoretical contributions.

read point-by-point responses

  1. Referee: [§2] §2 (hybrid model): the assumption that each pellet produces an instantaneous density jump interrupting continuous plasma dynamics is load-bearing for all subsequent stability analysis and the derived parameter constraints. The manuscript supplies only closed-loop simulations inside this model and provides neither experimental tokamak pellet data nor error bounds quantifying the effect of finite ablation/deposition profiles on the stability margins.

    Authors: The instantaneous-jump assumption is central to the hybrid model and enables the derivation of explicit parameter constraints and practical stability guarantees. This modeling choice is standard in the pellet-fuelling literature because ablation and deposition timescales are orders of magnitude faster than the continuous plasma transport dynamics. We will revise §2 to include an expanded discussion of the assumption's validity, supported by citations to experimental tokamak studies on finite ablation profiles, and to state explicitly that the derived tuning rules are intended as nominal guidelines that may require empirical adjustment. The work remains theoretical and simulation-based; we cannot supply new experimental data or quantitative error bounds derived from tokamak measurements. revision: partial

  2. Referee: [§5] §5 (stability analysis): the practical stability guarantees and explicit actuator/controller constraints are obtained under the hybrid model; without sensitivity analysis to model mismatch (e.g., non-instantaneous jumps), it remains unclear whether the tuning rules remain valid when the continuous inter-pellet evolution deviates from the assumed dynamics.

    Authors: The formal results are derived under the stated hybrid model. In the revised manuscript we will add a sensitivity study in §5 (or a new subsection) that replaces instantaneous jumps with smoothed, finite-duration deposition profiles and verifies that the proposed actuator and controller parameter constraints continue to produce bounded density responses in simulation. While this does not yield new formal guarantees for the perturbed system, it provides concrete numerical evidence of robustness to the most relevant model mismatch. revision: yes

standing simulated objections not resolved
  • Supplying original experimental tokamak pellet data or quantitative error bounds on finite ablation effects, as the study is confined to theoretical modeling and numerical simulations.

Circularity Check

0 steps flagged

No circularity: stability constraints derived from explicit hybrid model analysis

full rationale

The paper proposes a hybrid model capturing instantaneous pellet-induced density jumps interrupting continuous plasma dynamics, then performs formal analysis on this model and on both the neuromorphic and sigma-delta controllers to obtain explicit parameter constraints that guarantee practical stability. No step reduces the claimed result to a fitted input, self-definition, or load-bearing self-citation; the guarantees follow directly from the stated model assumptions and controller definitions rather than being tautological with the inputs. The derivation chain is therefore self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit list of free parameters or axioms; the hybrid model itself is treated as given, and stability guarantees rest on unstated assumptions about pellet-induced density jumps and plasma time scales.

pith-pipeline@v0.9.0 · 5555 in / 1070 out tokens · 24118 ms · 2026-05-07T11:24:42.854985+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    Bibitem after note You are using a bibitem after a note in a subbibitems environment; note should the last item in a subbibitems environment \@itemnote @bb@error [] @noitemnote \@tempa \@noitemnote \@noitemnote \@itemnote @bibitem\@bibitem @lbibitem\@lbibitem \@bibitem#1 \@itemnote @bibitem #1 \@tempa @noitemnote \@lbibitem[#1]#2 \@itemnote @lbibitem[#1] ...

    work page

  2. [2]

    @stdbsttrue @ctr \@lbibitem [ @ctr] \@bibitem##1 @ctr \@lbibitem[ @ctr] ##1 @bb@error\@mkbberr @filesw @natbibloaded \@auxout \@itemslabel @bibnum a-- @ctr \@auxout \@itemslabel :s @bibnum \@auxout \@itemslabel @bibnum a-- @ctr \@auxout \@itemslabel :s @bibnum @ctr @bibnum @citex\@citex \@tempcntc @citex[#1]#2 @filesw \@auxout #2 \@tempcnta @\@tempcntb @n...

    work page 1991

  3. [3]

    Blanken, T.C., Felici, F., Rapson, C.J., de Baar, M.R., & Heemels, W.P.M.H. (2018). Control-oriented modeling of the plasma particle density in tokamaks and application to real-time density profile reconstruction. Fusion Eng. Des. , 126, 87-103

    work page 2018

  4. [4]

    Bosman, T.O.S.J., van Berkel, M., & de Baar, M.R. (2021). Model-based electron density profile estimation and control, applied to ITER. J. Phys. Commun. , 5(11), 115015

    work page 2021

  5. [5]

    Bosman, T.O.S.J., Koechl, F., Ho, A., de Baar, M.R., Krishnamoorthy, D., & van Berkel, M. (2023). Integrated model control simulations of the electron density profile and the implications of using multiple discrete pellet injectors for control. Nucl. Fusion , 63, 126047

    work page 2023

  6. [6]

    Combs, S.K., Milora, S.L., Baylor, L.R., Foust, C.R., Gethers, F.E., & Sparks, D.O. (1988). A three-barrel repeating pneumatic pellet injector for plasma fueling of the Joint European Torus. J. Vac. Sci. Technol. A , 6, 1901-1904

    work page 1988

  7. [7]

    Derks, G.L., Kool, B., Vincent, C., Elmore, S., Henderson, S.S., Koenders, et al. (2024). Development of real-time density feedback control on MAST-U in L-mode. Fusion Eng. Des. , 202, 114387

    work page 2024

  8. [8]

    Fleschner, F. (2025). Wendelstein 7-X sets new performance records in fusion research. IPP. https://www.ipp.mpg.de/5532945/w7x

    work page arXiv 2025

  9. [9]

    Gerstner, W., & Kistler, W.M. (2002). Spiking neuron models: Single neurons, populations, plasticity . Cambridge University Press, Cambridge

    work page 2002

  10. [10]

    Geulin, E., & P\'egouri\'e, B. (2022). Pellet core fueling in tokamaks, stellarators and reversed field pinches. Plasma and Fusion Research , 17, 2102101

    work page 2022

  11. [11]

    Goebel, R., Sanfelice, R.G., & Teel, A.R. (2012). Hybrid Dynamical Systems: Modeling, Stability, and Robustness . Princeton University Press, Princeton, NJ

    work page 2012

  12. [12]

    Heemels, W.P.M.H., De Schutter, B., Lunze, J., & Lazar, M. (2010). Stability analysis and controller synthesis for hybrid dynamical systems. Philos. trans., Math. phys. eng. sci , 368(1930), 4937-4960

    work page 2010

  13. [13]

    Inose, H., Yasuda, Y., & Murakami, J. (1962). A Telemetering System by Code Modulation - - Modulation. IRE Transactions on Space Electronics and Telemetry , 8(3), 204-209

    work page 1962

  14. [14]

    Jansen, L.L.T.C., Petri, E., van Berkel, M., & Heemels, W.P.M.H. (2025). Nuclear fusion plasma fuelling with ice pellets using a neuromorphic controller. 2025 IEEE 64th Conference on Decision and Control (CDC) , 45-50

    work page 2025

  15. [15]

    Kudlacek, O., David, P., Gomez, I., Gr\"ater, A., Sieglin, B., Treutterer, W., et al. (2024). Overview of advances in ASDEX Upgrade plasma control to support critical physics research for ITER and beyond. Nucl. Fusion , 64, 056012

    work page 2024

  16. [16]

    Medvedeva, T., Casta\ nos, F., & Franci, A. (2025). Formalizing Neuromorphic Control Systems: A General Proposal and a Rhythmic Case Study. 2025 IEEE 64th Conference on Decision and Control (CDC) , 962-967

    work page 2025

  17. [17]

    Lang, P.T., Blanken, T.C., Dunne, M., McDermott, R.M., Wolfrum, E., Bobkov, V., et al. (2018). Feedback controlled, reactor relevant, high-density, high-confinement scenarios at ASDEX Upgrade. Nucl. Fusion , 58, 036001

    work page 2018

  18. [18]

    Lapicque, L. (1907). Recherches quantitatives sur l’excitation electrique des nerfs traitee comme une polarization. Journal de Physiologie et de Pathologie G\'en\'erale , 9, 620-635

    work page 1907

  19. [19]

    Lennholm, M., Budd, T., Felton, R., Gadeberg, M., Goodyear, A., Milani, F., & Sartori, F. (2000). Plasma control at JET. Fusion Eng. Des. , 48(1-2), 37-45

    work page 2000

  20. [20]

    Orrico, C.A., van Berkel, M., Bosman, T.O.S.J., Heemels, W.P.M.H., & Krishnamoorthy, D. (2023). Mixed-Integer MPC Strategies for Fueling and Density Control in Fusion Tokamaks. IEEE Control Syst. Lett. , 7, 1897-1902

    work page 2023

  21. [21]

    Petri, E., Scheres, K.J.A., Steur, E., & Heemels, W.P.M.H. (2024). Analysis of a Simple Neuromorphic Controller for Linear Systems: A Hybrid Systems Perspective. 2024 IEEE 63rd Conference on Decision and Control (CDC) , 8578-8583

    work page 2024

  22. [22]

    Petri, E., Postoyan, R., & Heemels, W.P.M.H. (2025). Rhythmic neuromorphic control of a pendulum: A hybrid systems analysis. 2025 IEEE 64th Conference on Decision and Control (CDC) , 19-24

    work page 2025

  23. [23]

    Ploeckl, B., & Lang, P.T. (2013). The enhanced ASDEX Upgrade pellet centrifuge launcher. Rev. Sci. Instrum. , 84, 103509

    work page 2013

  24. [24]

    Ploeckl, B., Lang, P.T., Kircher, M., Bock, A., Gude, A., Janky, F., et al. (2021). Targeting a Versatile Actuator for EU-DEMO: Novel Control Scheme for Multisource Pellet Injector. Fusion Sci. Technol. , 77(3), 199-205

    work page 2021

  25. [25]

    Ravensbergen, T., de Vries, P.C., Felici, F., Blanken, T.C., Nouailletas, R., & Zabeo, L. (2018). Density control in ITER: an iterative learning control and robust control approach. Nucl. Fusion , 58, 016048

    work page 2018

  26. [26]

    Reiss, J. (2008). Understanding Sigma-Delta Modulation: The Solved and Unsolved Issues. AES: J. Audio Eng. Soc. , 56

    work page 2008

  27. [27]

    Ribar, L., & Sepulchre, R. (2021). Neuromorphic Control: Designing Multiscale Mixed-Feedback Systems. IEEE Control Syst. , 41(6), 34-63

    work page 2021

  28. [28]

    Romanelli, M., Parail, V., da Silva Aresta Belo, P., Corrigan, G., Garzotti, L., Harting, D., et al. (2015). Modelling of plasma performance and transient density behaviour in the H-mode access for ITER gas fuelled scenarios. Nucl. Fusion , 55, 093008

    work page 2015