An Electronic Nose for Recognizing Combustible Gases using Thick Film Sensor Array and Neural Network

摘要

We propose an electronic nose for recognizing combustible gases using a SnC>2 based thick film sensor array and neural network for pattern recognizer, within the ranges of TLVs and LELs. The electronic nose can not only identify the kinds of the combustible gases but also recognize concentration values of the identified gas. A sensor array with ten discrete sensors was developed with the aim of recognizing the kinds and quantities of combustible gases, such as methane, propane, butane, and carbon monoxide within the standard ranges. The sensor array consisted of gas-sensing materials of SnO2, plus a heating element based on a meandered platinum layer. The sensors on a sensor array were designed to produce a uniform thermal distribution and show a high and broad sensitivity to low concentrations of about 100 ppm. Using the sensing signals of the array along with an artificial neural network, an electronic nose system was then implemented. The characteristics of the multi-dimensional sensor signals obtained from sensors were analyzed using the principal component analysis (PCA), and a gas pattern recognizer was implemented using a neural network with an error-back-propagation learning algorithm. The simulation and experimental results demonstrated that the proposed electronic nose was effective in identifying combustible gases. For real time processing, a digital signal processor (DSP) board was then used to implement the electronic nose.