Optimization of Sensor Arrays for the Identification of Abalone Flavoring Liquids

Abstract

To optimize the sensor arrays which are used to classify and identify these fluids. It preliminarily screens the response performance of single sensors in the testing system using the intra-class mean square, the F value and the P value by one-way analysis of variance (one-way ANOVA) method, and then performs significance analysis of the screened sensors according to the multiple comparison analysis method, classify them by significance and then group them into three different sensor arrays. After that, this paper performs principal component analysis and cluster analysis on the signals of each sensor array, and the results show that the optimized sensor arrays all have better performance than the ones before optimization in classification and identification of samples, and that in particular, the optimized array I consisting of Au, Pt, Pd and W has the best performance and can be applied in the classification and identification of abalone flavoring liquids.