Wireless Sensor Network-based Detection of Poisonous Gases Using Principal Component Analysis

This work utilizes a statistical approach of Principal Component Analysis (PCA) towards the detection Methane (CH4)-Carbon Monoxide (CO) Poisoning occurring in coal mines, forest fires, drainage systems etc. where CH4 and CO emissions are very high closed buildings or confined spaces during oxidation processes. Both methane carbon monoxide highly toxic, colorless odorless gases. gases have their own toxic levels to be detected. But combined presence, toxicity either one goes unidentified may due low which lead an explosion. By using PCA, correlation data is carried out by identifying areas (along principal component axis) explosion suppression action can triggered earlier thus avoiding adverse effects massive explosions. Wireless Sensor Network deployed simulations with heterogeneous sensors (Carbon sensors) NS-2 Mannasim framework. The rise value even when below level become hazardous people around. Thus our proposed methodology will detect presence both (CH4 CO) provide early warning order avoid any human losses effects.