P34 - Optimisation of Convolutional Neural Networks for MOS Gas Sensors

Event

17. Dresdner Sensor-Symposium 2024
2024-11-25 - 2024-11-27
Dresden

Band

Poster

Chapter

14. Smart Sensors/Künstliche Intelligenz in der Sensorik

Author(s)

J. Petry, D. Schu, T. Mertin, D. Arendes - Universität des Saarlandes,Saarbrücken/D

Pages

217 - 222

DOI

10.5162/17dss2024/P34

ISBN

978-3-910600-04-1

Price

free

Abstract

Metal oxide semiconductor (MOS) gas sensors are widely used in applications fields ranging from indoor air quality (IAQ) monitoring and control [1] to fire detection [2]. These sensors offer a low-cost in situ alternative to gas chromatography [3]. Temperature-cycled operating (TCO) modes are proven to enhance selectivity as well as sensitivity [4]. However, MOS sensors still require calibration against a known reference, which is often provided by a gas mixing apparatus [5]. Using the measured conductivity of the sensors exposed to a unique gas mixture (UGM), machine learning (ML) algorithms can be trained to classify or quantify gases.