CoilStellaration: A Dataset and Benchmark for Engineering-Aware Stellarator Coilset Generation
Organizations: Proxima Fusion, Munich, Germany
Abstract
Stellarators are fusion devices that use externally generated, three-dimensional magnetic fields to confine hot plasma and offer a promising route to commercial fusion. Their electromagnetic coils pose a challenging design problem: they must reproduce the target magnetic field while satisfying engineering constraints and remaining practical and cost-effective to manufacture. Finding suitable geometries is a highly nonlinear inverse problem that generally has neither a closed-form nor a unique solution. This problem lies at the heart of stellarator design. We introduce CoilStellaration, a dataset and benchmark for learning the mapping between coil geometries from target plasma magnetic fields and engineering requirements. The dataset contains 177,000 optimized filamentary coilsets targeting plasma equilibria from ConStellaration, with sampled engineering requirements and achieved geometric and magnetic-field metrics. It spans multiple coil counts and field-period numbers, supporting studies of both specialized and shared models across configurations. The benchmark asks models to predict coils conditioned on a target equilibrium and requirements on coil-coil and coil-plasma spacing, curvature, and magnetic-field error. We show that magnetic-field accuracy remains the main challenge, with learned models rarely satisfying all requirements simultaneously. However, using learned predictions to initialize conventional coil optimization can accelerate early progress. We release baseline models and evaluation code to support reproducible research on both direct coil generation and learned optimization initializations. By lowering the barrier for machine learning and optimization researchers to engage with stellarator coil design, we aim to accelerate progress toward practical, cost-effective stellarator power plants.
Figures & tables
Appendix figures & tables15 assets
Supplementary material from the paper’s appendix.
Appendix
| Property | ConStellaration | QUASR | CoilStellaration |
|---|---|---|---|
| Plasma boundaries | 14,496 target boundaries | ||
| Coilsets | — | 266,549 REGCOIL initializations; 177,341 completed DESC coilsets | |
| Coilsets per boundary | — | One paired coilset per entry | On average, 18.4 REGCOIL initializations and 12.2 DESC coilsets |
| Configuration family | QI-like | QA and QH | QI-like |
| Number of field periods | 1–5 | 1–6 | 2–5 |
| Unique coils per half field period | — | 1–13 | 4–8 |
| 1 iteration | 500 iterations | |||
|---|---|---|---|---|
| Requirement | Per-coil ensemble | REGCOIL | Per-coil ensemble | REGCOIL |
| 38% | 92% | 69% | 80% | |
| 58% | 10% | 72% | 52% | |
| 90% | 94% | 84% | 84% | |
| 0% | 0% | 65% | 47% | |
| All | 0% | 0% | 27% | 18% |
| Stage | Criterion | Physical / Numerical Rationale |
|---|---|---|
| Upstream Convergence | Discard configurations with missing attributes caused by non-convergence in stage-1 (optimization) or stage-2 (forward models). | |
| Optimization Provenance | Require the boundary to originate from active VMEC++ or standard VMEC omnigenity optimization, rejecting unoptimized initial guesses. | |
| Positive Orientation | Enforce to standardize spatial orientation for downstream control-coil synthesis and neural network ingestion. | |
| Metric Validity | Require a finite Goodman QI residual ( ). | |
| Field Period Bounds | Restrict this release to . | |
| Equilibrium Mapping | Exclude boundaries that fail to instantiate or converge to a force-balanced state within the VMEC++ computational framework. |
| Reference Design | (coil-to-coil distance) | (coil-to-plasma distance) | (coil curvature) |
|---|---|---|---|
| W7-X [ 1 ] | 0.53 | 0.70 | 1.31 |
| HSX [ 27 ] | 0.76 | 1.18 | 1.4 |
| NCSX [ 18 ] | 0.38 | 0.67 | 3.18 |
| Stellaris [ 17 ] | 0.51 | 1.05 | 2.03 |
| Infinity Two [ 10 ] | 0.52 | 0.97 | 2.4 |
| Sampled Range | 0.3 - 0.7 | 0.5 - 1.0 | 1.5 - 5.0 |
| REGCOIL | (%) | DESC | % of DESC | DESC / REGCOIL (%) | |
|---|---|---|---|---|---|
| 2 | 30 481 | 11.4 | 30 234 | 17.0 | 99.2 |
| 3 | 104 075 | 39.0 | 89 503 | 50.5 | 86.0 |
| 4 | 63 609 | 23.9 | 33 077 | 18.7 | 52.0 |
| 5 | 68 384 | 25.7 | 24 527 | 13.8 | 35.9 |
| Total | 266 549 | 100.0 | 177 341 | 100.0 | 66.5 |
| Stage | All | ||||
|---|---|---|---|---|---|
| Released samples | 266 549 | 30 481 | 104 075 | 63 609 | 68 384 |
| Successful optimization | 177 341 | 30 234 | 89 503 | 33 077 | 24 527 |
| Full pipeline executed | 163 908 | 30 221 | 78 906 | 33 020 | 21 761 |
| + Coil-plasma error | 140 965 | 26 218 | 66 237 | 29 368 | 19 142 |
| + Coil-coil error | 138 954 | 25 692 | 65 554 | 28 949 | 18 759 |
| + Curvature error | 138 765 | 25 614 | 65 455 | 28 939 | 18 757 |
| Metric | Direction | |
|---|---|---|
| (coil-to-coil distance) | lower bound | |
| (coil-to-plasma distance) | lower bound | |
| (coil curvature) | upper bound | |
| (field error) | upper bound |
| Architecture | ||||
|---|---|---|---|---|
| MLP | ||||
| ResMLP | ||||
| EnsembleMLP | ||||
| EnsembleResMLP |
| Reference metric | with strict feasibility | with score |
|---|---|---|
| Uniform scale | |||||
|---|---|---|---|---|---|
| vs nominal | — | ||||
| Field-error | |||||
| vs nominal | — |
| Parameter | Value | Description |
| 512 | Width of the primary MLP trunk. | |
| 4 | Four hidden stages; separate output projection. | |
| 64 | Hidden dimension of the specialized currents head. | |
| Dropout | 0.2 | Dropout rate (exclusive to Residual models). |
| 5 | Number of members in ensemble variants. | |
| 1.2 | Spectral-scaling exponent for bidirectional flat mapping. |
| Parameter | Value |
|---|---|
| Seeds | 0, 1, 2 |
| Training budget | Up to 600,000 updates with 1,000-step warmup. |
| Batch / selection | Train batch 16, evaluation batch 256, and evaluation every 500 updates. |
| Stopping Patience | 100,000 updates with best checkpoint retained. |
| Optimizer | AdamW with learning rate schedule to , and weight decay , and global norm clip 1.0. |
Inverse Design
Inverse design maps desired properties or behaviors to materials, geometries, molecules, and systems that can realize them.
See all 118 papers