Answers are grounded in this paper and cite their sources.
To overcome the trade-off between optical path length (OPL) and physical volume (V) in conventional multi-pass cell (MPC) design, and to address the issues of uncontrollable spot patterns and susceptibility to interference fringes in dense spot configurations, this work proposes a novel automatic design paradigm of multi-objective optimization for ultra-compact plane-spherical MPC by synergizing the non-dominated sorting genetic algorithm III (NSGA-III) with K-Means clustering for the first time. In this method, K-means clustering is transformed from a mere post-processing tool into an intrinsic objective function that quantifies the regularity of the spot distribution, which is then fed into the NSGA-III engine. This enables the simultaneous multi-objective optimization of OPL, OPL-to-volume ratio (RLV), spot uniformity, and inter-spot safety spacing. Using this autonomous design approach, an ultra-compact MPC with a recorded RLV of 48.18 cm−2 is achieved, enabling automatically controllable dense spot patterns while effectively avoiding optical crosstalk and overlap. Experimental results demonstrate excellent agreement between the measured and simulated spot patterns for three designed spot configurations (concentric two-rings, independent 17-rings, and independent nine-rings). An acetylene (C2H2) light-induced thermoelastic spectroscopy (LITES) sensor with the independent 17-ring MPC (volume of only ∼16 mL) was constructed for performance validation, achieving a minimum detection limit of 6.1 ppb and exhibiting a concentration linearity of 0.99. This demonstrates that the proposed auto-designed MPC method, based on the synergy of NSGA-III and K-means clustering, realizes a breakthrough integration of ultra-high sensitivity and field-ready miniaturization for gas sensors.
No takes yet. Share an insight, caveat, or question.
Sun et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: