SEMICONDUCTOR GAS SENSOR ARRAY-BASED INTELLIGENT SYSTEM FOR COMBUSTION GAS ANALYSIS

摘要

Gas sensor arrays have been fabricated using semiconducting gas sensors obtained from a commercial source, as well as those produced in-house. An intelligent system for the identification of combustion gases such as methane, propane, ethanol, carbon monoxide and their mixtures was assembled by interfacing gas sensor arrays to a microcomputer and pattern recognition software, i.e., Principal Component Analysis (PCA). PCA was used to analyse the electrical response of the gas sensor arrays exposed to various combustion gases and their mixtures. Preliminary results are very encouraging. The data clearly indicates that principal component analysis can be successfully used to qualitatively identify or classify these gases and their mixtures using semiconducting gas sensor arrays.