研究目的
To develop a radiative transfer model for FT-IR spectroscopy to create close-to-reality toxic gas spectra and use these spectra to train a support vector machine (SVM) for chemical gas detection, overcoming the limitations of measuring toxic gases in open environments.
研究成果
The proposed radiative transfer model and SVM algorithm successfully generated close-to-reality toxic gas spectra and improved the detection performance of FT-IR spectroscopy for chemical warfare agents. The SVM algorithm demonstrated a high probability of accurate detection while reducing false negative and false-positive detection rates in both indoor and outdoor experiments.
研究不足
The use of certain toxic gases is prohibited, requiring the use of simulant gases for outdoor measurements. The study acknowledges the limitations of obtaining toxic gas spectra in outdoor environments and the need for secured facilities and safety equipment for such experiments.
1：Experimental Design and Method Selection:
The study is based on infrared radiative transfer models for passive FT-IR spectroscopy, improved using the atmospheric transmittance obtained from MODTRAN and considering the spectral responses of the detectors.
2：Sample Selection and Data Sources:
Nerve agents (tabun, sarin, soman, and cyclosarin) and a simulant gas (sulfur hexafluoride) were used for indoor and outdoor experiments.
3：List of Experimental Equipment and Materials:
Two off-the-shelf FT-IR gas detection systems (Porthos and Hi-90) were used.
4：Experimental Procedures and Operational Workflow:
The radiance spectrum measured by FT-IR sensors was approximated based on a three-layer model consisting of background, a vapor cloud, and the atmosphere. The emulated gas spectra were used to train SVM for chemical gas detection.
5：Data Analysis Methods:
The detection performance was evaluated using the SVM algorithm, and its performance was compared with Pearson’s cross-correlation and adaptive subspace detector.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
