Drift Correction of Chemical Sensors

Abstract

Artificial olfaction systems that try to mimic human olfaction by using arrays of gas chemical sensors combined with pattern recognition methods represent a potentially economic tool in many areas of industry such as: perfumery, food and drinks production, clinical diagnosis, health and safety, environmental monitoring and process control. However, successful applications of these systems are still largely limited to specialized laboratories. Among others, sensor drift, the lack of stability over time still limit real industrial setups. This chapter presents and discusses an evolutionary based adaptive drift-correction method designed to work with state-of-the-art classification algorithms. The proposed system exploits a leading-edge evolutionary strategy to iteratively tweak the coefficients of a linear transformation able to transparently transform raw sensors measures in order to mitigate negative effects of the drift. The optimal correction strategy is learned without a-priori models or other hypothesis on the behavior of physical-chemical sensors. Preliminary results have been published in [49].