Retrieval of gas concentrations in optical spectroscopy with deep learning

A novel direct absorption spectroscopy gas sensing system based on end-to-end deep neural networks was proposed for measurements of gas concentration. One-dimensional convolutional neural network and deep multi-layer perceptron were explored to measure the concentrations of methane and acetylene. The accurate measurement results for both gases demonstrated that deep neural networks based direct absorption spectros-copy technique can be reliably applied to different gas molecules. The developed gas sensing system achieved more precise concentration retrieval compared with that of wavelength modulation spectroscopy, and fast computation speed as well as robustness to noisy conditions, laser aging and circuit parameter variation simultaneously. The combination of deep neural networks and direct absorption spectroscopy provides new ideas for the further research of gas absorption spectroscopy.

Automated summary

A novel gas sensing system based on end-to-end deep neural networks for measuring gas concentration was proposed. The system utilized one-dimensional convolutional neural networks and deep multi-layer perceptron to measure concentrations of methane and acetylene, achieving accurate measurement results and demonstrating reliability for different gas molecules. The combination of deep neural networks and direct absorption spectroscopy provided new ideas for further research in gas absorption spectroscopy, showing promising potential for precise concentration retrieval in noisy conditions.