Scholarlink

<p>A parallel heterodyne light-induced thermoelastic spectroscopy (PH-LITES) sensor is proposed for high-speed and high-sensitivity multi-gas detection for the first time. Within the collaborative signal enhancement architecture (CSEA), high sensitivity and high-speed detection are achieved at the physical layer. A self-designed cylindrical multi-pass cell (MPC) with a recorded high optical path length to volume ratio (OPL/V = 37.4 cm<sup>–2</sup>) and a four‑tine quartz tuning fork (QTF) with a low resonant frequency (<i>f</i><sub>0</sub> = ~7.9 kHz) work synergistically to enhance detection responsivity, establishing a robust foundation for highly sensitive detection of low-concentration gas mixtures. High‑speed capability is enabled by parallel heterodyne modulation, where a single QTF is excited to generate a composite transient response signal, allowing for the rapid, simultaneous acquisition of spectral information from multiple gases. At the information layer, intelligent processing is implemented via a collaborative intelligent processing architecture (CIPA) integrating convolutional neural networks (CNN), a hybrid attention mechanism (HAM), and bidirectional long short-term memory (BiLSTM). The CNN-HAM-BiLSTM architecture performs feature extraction, attention-based enhancement, and temporal modeling to enable accurate concentration retrieval from the parallel spectra derived from a single QTF output. Experimental validation using methane (CH<sub>4</sub>) and acetylene (C<sub>2</sub>H<sub>2</sub>) achieved minimum detection limits (MDLs) of 378 ppb and 285 ppb, respectively, within a 4 s scanning time. The proposed system offers an efficient solution for applications requiring rapid and sensitive multi-gas detection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

High-speed and high-sensitivity multi-gas detection based on parallel heterodyne LITES sensor

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月29日

查看全文
  • Haiyue Sun,
  • Shunda Qiao,
  • Ying He,
  • Yuanzhi Wang,
  • Jinfeng Hou,
  • Chu Zhang,
  • Yongkang Dong,
  • Yufei Ma

摘要

A parallel heterodyne light-induced thermoelastic spectroscopy (PH-LITES) sensor is proposed for high-speed and high-sensitivity multi-gas detection for the first time. Within the collaborative signal enhancement architecture (CSEA), high sensitivity and high-speed detection are achieved at the physical layer. A self-designed cylindrical multi-pass cell (MPC) with a recorded high optical path length to volume ratio (OPL/V = 37.4 cm–2) and a four‑tine quartz tuning fork (QTF) with a low resonant frequency (f0 = ~7.9 kHz) work synergistically to enhance detection responsivity, establishing a robust foundation for highly sensitive detection of low-concentration gas mixtures. High‑speed capability is enabled by parallel heterodyne modulation, where a single QTF is excited to generate a composite transient response signal, allowing for the rapid, simultaneous acquisition of spectral information from multiple gases. At the information layer, intelligent processing is implemented via a collaborative intelligent processing architecture (CIPA) integrating convolutional neural networks (CNN), a hybrid attention mechanism (HAM), and bidirectional long short-term memory (BiLSTM). The CNN-HAM-BiLSTM architecture performs feature extraction, attention-based enhancement, and temporal modeling to enable accurate concentration retrieval from the parallel spectra derived from a single QTF output. Experimental validation using methane (CH4) and acetylene (C2H2) achieved minimum detection limits (MDLs) of 378 ppb and 285 ppb, respectively, within a 4 s scanning time. The proposed system offers an efficient solution for applications requiring rapid and sensitive multi-gas detection.