Simultaneous Spectrophotometric Determination of O-nitro-aniline, M-nitro-aniline and P-nitro-aniline Using a Wavelet-based Generalized Regression Neural Network

This paper suggests a novel method named WTGRNN, which is based on generalized regression neural network (GRNN) with wavelet transform (WT) as a pre-processing tool for the simultaneous spectrophotometric determination of o-nitro-aniline, m-nitro-aniline and p-nitro-aniline. Wavelet representations of signals provide a local time-frequency description, thus in the wavelet domain, the quality of noise removal can be improved. GRNN was applied for overcoming the convergence problem met in back propagation training and facilitating nonlinear calculation. In this case, by optimization, wavelet function, decomposition level and the width (sigma) of GRNN for WTGRNN were selected as Coiflet 1, 5 and 0.4 respectively. The relative standard errors of prediction for all components with WTGRNN and GRNN were 4.93% and 6.56% respectively. The proposed method has been successfully applied to analyze overlapping spectra and was proven to be better than GRNN.