Advanced Gas Sensor Design: A Machine Learning Perspective on Photonic Crystals

A novel machine learning (ML) based approach is introduced for designing photonic crystal (PhC) gas sensors to detect SF6, CH4, and CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>. In this work, we have implemented both the algorithm Support Vector Machine (SVM) and Random Forest (RF) then comparing both algorithms, achieved the highest accuracy of 97.01% for SVM in case of gas classification. We have designed the sensor integrating machine learning models for real-time classification. We have run extensive simulations that provide dataset over a wide range of temperatures and pressures. Using this dataset, we have trained and optimized SVM and RF classifiers. The accuracies of gas detection get improved with machine learning along with simple sensor designs, which then presents an effective tool for density prediction. This work has opened pathways toward developing efficient, compact, and highly accurate gas sensors for environmental applications.