ONLINE AUTOMATED ANALYTICAL SIGNAL DIAGNOSIS IN SEQUENTIAL INJECTION-ANALYSIS SYSTEMS USING ARTIFICIAL NEURAL NETWORKS

This paper describes an automated analytical system able to diagnose m ultivariate spectrophotometric responses, with the aim of detecting fa ulty responses and assigning causes to the symptoms detected. Not only does this system detect faulty spectra, but it is also capable of mod ifying, by means of a 'feed-back response', the entire analytical syst em, and, when it is necessary, to report the conditions of the sequent ial injection analysis system to give an on-line diagnosis signal. Art ificial neural networks (ANNs), in particular counter-propagation neur al networks, have been applied to detect faults and diagnose signals o btained in a sequential injection analysis system. This strategy has b een used to analyse natural water samples and, in particular, to simul taneously determine calcium and magnesium by means of spectrophotometr ic detection of the complex which both cations form with the reagent A rsenazo(III).