[Application of artificial neural network to simultaneous spectro-fluorimetric determination of phenol and resorcinol].

By the means of artificial neural network and Levenberg-Marquardt back-propagation train algorithm, the two components of phenol, resorcinol were determined simultaneously, in which the fluorescence spectra overlapped. In 290-345 nm, the fluorescence intensity at 14 wavelengths were taken as character of artificial neural network, and samples were arranged by method of equality design. The mean recovery of phenol and resorcinol were 100.2%, 99.99% respectively. The RSD of the results were 0.4% and 1.3%.