Artificial neural networks (ANNs) in the analysis of polycyclic aromatic hydrocarbons in water samples by synchronous fluorescence Original Research Article

Backpropagation artificial neural networks, principal component regression and partial least squares have been compared in order to establish the best multivariate calibration models for the analysis of mixtures of polycyclic aromatic hydrocarbons containing 10 of these compounds (anthracene, benz[a]anthracene, benzo[a]pyrene, chrysene, fluoranthene, fluorene, naphthalene, perylene, phenanthrene and pyrene). The synchronous fluorescence spectra (recorded at wavelength increments of 50 and 100 nm) of 85 standards, with concentrations ranging from 0 to 20 ng ml−1, have been used for this purpose.