Nonlinear PCA based micro gas sensor array signal processing

A nonlinear superposition model was proposed based on the common linear additive model the micro gas sensor array signal processing to improve the precision of quantification and identification. According to the nonlinear model, the Nonlinear Principal Component Analysis (NLPCA) was proposed to process the response signals obtained from a 4 Micro-hotplate (MHP) based gas sensor array. Com-pared with the analyzing results obtained from Principal Component Analysis (PCA), which bases onthe linear additive model, the accuracy of gas component identification and concentration quantification are improved greatly.