MAGNA-FLOW Documentation

📖 Overview

"The magnetic field does not merely confine the plasma — it is the plasma's memory. MAGNA-FLOW reads that memory and rewrites the future before the instability can."

MAGNA-FLOW introduces the first physics-informed AI framework for real-time magnetohydrodynamic dissipation control in high-conductivity turbulent plasma systems. Built on three orthogonal neural constructs spanning spectral operator learning, physics-constrained optimization, and model-predictive control, MAGNA-FLOW achieves unprecedented efficiency in suppressing edge-localized modes in tokamaks, breathing-mode oscillations in Hall thrusters, and Hartmann turbulence in liquid-metal nuclear reactors.

🏗️ 3-Core Architecture

MAGNA-FLOW operates on the 6-component MHD state vector v(r,t) = (u_x, u_y, u_z, B_x, B_y, B_z) through three synergistic neural constructs:

M-FNO — Magnetic Fourier Neural Operator

Generalizes the Fourier Neural Operator to the full coupled MHD setting with learnable 6×6 complex spectral kernels capturing the complete Alfvénic coupling tensor. The Helmholtz-Hodge divergence-free projector ensures div(u) = div(B) = 0 as a hard architectural guarantee.

from magna_flow import MFNO

mfno = MFNO(
    spatial_dim=256,
    k_max=64,
    channels=256,
    layers=8,
    complex_kernel=6x6
)

H-PINN — Hydromagnetic Physics-Informed Network

Enforces complete MHD conservation laws — momentum, induction, solenoidal constraint, magnetic helicity evolution, and Onsager cross-coupling symmetry — as hard loss terms with NTK rebalancing and causal temporal weighting.

L-Flux — Lorentz Flux Resolver

Model-predictive control engine tracking the Maxwell stress tensor T_M in real time. When minimum eigenvalue λ_min approaches zero (magnetic pressure collapse), L-Flux pre-emptively actuates correction fields with 312 µs lead time.

📐 Core Equations

v(r,t+dt) = W·v(r,t) + F⁻¹[ R_φ(k)·F[v](k) ]

∂B/∂t = ∇ × (u × B) + η·∇²B

T_M^{ij} = (1/μ₀)·[B_i B_j − (1/2)δ_{ij}|B|²]

min∫∫[σ_Ohm + σ_visc] dr dt
   s.t. λ_min(T_M) ≥ λ_safe, |B_ctrl| ≤ B_max

📦 Installation

# From PyPI (stable)
pip install magna-flow-engine

# From source
git clone https://gitlab.com/gitdeeper11/MAGNA-FLOW.git
cd MAGNA-FLOW && pip install -e .

# With CUDA acceleration
pip install magna-flow-engine[cuda]

🔧 API Reference

MHDStateTracker

from magna_flow import MHDStateTracker

tracker = MHDStateTracker(
    spatial_dim=256,
    k_max=64,
    fluid='plasma_deuterium',
    enforce_helicity=True,
    lflux_horizon_us=500
)

tracker.load_weights('experiments/weights/')
tracker.step(dt=1e-6, env_obs=observation)
eta = tracker.get_efficiency_index()

🧩 Core Modules

📊 Validation Summary

👤 Author

📝 Citation

"The magnetic field does not merely confine the plasma — it is the plasma's memory. MAGNA-FLOW reads that memory and rewrites the future before the instability can."