Drift compensation of gas sensor array by matrix transform and genetic algorithm based on database

A new drift compensation method based on matrix transform and genetic algorithm (GA) is proposed in this article. Continuous experiments for a period of 10 months were carried out to establish a database, which was then used to evaluate our method. It can be found that the new method can considerably reduce adverse effects of drift and other factors on sensor arrays in electronic nose system (E-nose), thereby increasing the recognition accuracy accordingly. Also, it compares favorably with two existing drift compensation methods: component correction principal component analysis (CC-PCA) and orthogonal signal correction (OSC) based methods. Moreover, it does not require specific reference gases.