Automated identification of quasi-BIC modes in disordered resonator chains using spatial field envelopes and resonance descriptors
Abstract
Structural disorder turns bound states in the continuum (BICs) into quasi-BIC resonances and mixes them with spectrally close eigenmodes, so the target mode must be identified anew in every disorder realization. We develop an automated procedure for this task and test it on a chain of 50 ceramic disks, where disorder reorders the standing-wave eigenmodes at the edge of the BIC band: at disorder amplitude σ=0.05, the highest-quality-factor (Q) mode is the target in only 45\% of realizations, and above σ≈0.06 the highest-frequency mode is no longer the target. The target is identified from the spatial envelope of its field and its resonance parameters by an ensemble of a convolutional--recurrent neural network and two gradient-boosted decision-tree models. On 299 held-out realizations, the ensemble selects the target in 82.6\% of cases, against 60.9\% and 49.5\% for the highest-frequency and highest-Q rules, and in 69.5\% of cases for σ≥0.09, where the best of these rules reaches 57.6\%. The mean Q of the selected modes follows that of the manually labeled modes (⟨ Q⟩∝σk, k=−0.80±0.05 versus −0.86±0.04). After adaptation on 50 labeled realizations, the workflow identifies the quasi-BIC of a Ge2Sb2Te5 bar metasurface in 98--100\% of labeled cases, including a lossy phase where Q does not distinguish the mode.
Related
- Unifying Plasticity in Ordered and Disordered Matter using Topological and Geometrical Descriptors
- VoroTypes: automated enumeration of Voronoi cell topologies for structural identification of complex crystals in atomistic simulations
- 4DMulti: automated multicomponent identification at complex material interfaces
- An end-to-end experimental pipeline for automated track identification for paleo-detectors
- An Automated and Reproducible Workflow for Crack Identification and Damage Assessment of Fusion Materials
BibTeX
@misc{porvatov2026automated,
title = {{Automated identification of quasi-BIC modes in disordered resonator chains using spatial field envelopes and resonance descriptors}},
author = {Vadim A. Porvatov and Kliment V. Semushev and Ravshanjon Kh. Nazarov and Nikolai A. Vlasov and Alexander I. Solomonov and Zarina F. Kondratenko and Ekaterina E. Maslova},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.37967},
eprint = {2610.37967},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.37967},
}
RIS
TY - UNPB
TI - Automated identification of quasi-BIC modes in disordered resonator chains using spatial field envelopes and resonance descriptors
AU - Vadim A. Porvatov
AU - Kliment V. Semushev
AU - Ravshanjon Kh. Nazarov
AU - Nikolai A. Vlasov
AU - Alexander I. Solomonov
AU - Zarina F. Kondratenko
AU - Ekaterina E. Maslova
PY - 2026
DA - 2026/10/27
PB - arXiv
DO - 10.48550/arxiv.2610.37967
UR - https://arxiv.org/abs/2610.37967
AB - Structural disorder turns bound states in the continuum (BICs) into quasi-BIC resonances and mixes them with spectrally close eigenmodes, so the target mode must be identified anew in every disorder realization. We develop an automated procedure for this task and test it on a chain of 50 ceramic disks, where disorder reorders the standing-wave eigenmodes at the edge of the BIC band: at disorder amplitude σ=0.05, the highest-quality-factor (Q) mode is the target in only 45\% of realizations, and above σ≈0.06 the highest-frequency mode is no longer the target. The target is identified from the spatial envelope of its field and its resonance parameters by an ensemble of a convolutional--recurrent neural network and two gradient-boosted decision-tree models. On 299 held-out realizations, the ensemble selects the target in 82.6\% of cases, against 60.9\% and 49.5\% for the highest-frequency and highest-Q rules, and in 69.5\% of cases for σ≥0.09, where the best of these rules reaches 57.6\%. The mean Q of the selected modes follows that of the manually labeled modes (⟨ Q⟩∝σ^k, k=−0.80±0.05 versus −0.86±0.04). After adaptation on 50 labeled realizations, the workflow identifies the quasi-BIC of a Ge_2Sb_2Te_5 bar metasurface in 98--100\% of labeled cases, including a lossy phase where Q does not distinguish the mode.
ER -