Harmful gases detection using artificial neural networks of the environment

Abstract

This work describes a small, low-cost electronic nose device which can detect harmful substances that can harm human health, such as flammable gas like acetone, ethanol, butane as well as methane, among others. An artificial olfactory instrument consists of a set of metal oxide semiconductor sensors as well as a computer-based communications channel for signal gathering, proceeding, and presentation. We used three sensors instead of six, and the results were plotted as a variance, score as well as loading plot with crossvalidation. For gas identification, we use artificial neural network (ANN) and compare them to parallel factor analysis. Electronic nose (e-nose) has provided numerous benefits in a variety of logical study disciplines. Our goal is to create a sensor exhibit framework that can discriminate the most exceedingly contaminated gases while also being extremely responsive, precise, and less power consuming. Thus, for gas detection, we employ an ANN as well as make a comparison of results with parallel factor analysis (PARAFAC).