Sensor characterization for multisensor odor-discrimination system

Abstract

In recent years, with the advent of new and cheaper sensors, the use of olfactory systems in homes, industries, and hospitals has a new start. Multisensor systems can improve the ability to distinguish between complex mixtures of volatile substances. To develop multisensor systems that are accurate and reliable, it is important to take into account the anomalies that may arise because of electronic instabilities, types of sensors, and air flow. In this approach, 32 metal oxide semiconductor sensors of 7 different types and operating at different temperatures have been used to develop a multisensor olfactory system. Each type of sensor has been characterized to select the most suitable temperature combinations. In addition, a prechamber has been designed to ensure a good air flow from the sample to the sensing area. The multisensor system has been tested with good results to perform multidimensional information detection of two fruits, based on obtaining sensor matrix data, extracting three features parameters from each sensor curve and using these parameters as the input to a pattern recognition system. (C) 2012 Elsevier B.V. All rights reserved.