Automatic classification of polymer coating quality using artificial neural networks

Two artificial neural networks (ANN)-one for classification of polymer coat ing quality based on phase angle (Phi)-log frequency (f) data and one for c lassification based on log impedance modulus (/Z/)-log f data-have been tra ined using three sets of theoretical impedance spectra for polymer coated s teel-spectra for 'good', 'intermediate' and 'poor' coating quality. The tra ined ANNs have been tested using experimental impedance spectra for six dif ferent polymer coating systems on steel collected during exposure at a remo te marine test site for exposure periods up to one year. In general, excell ent agreement between the predictions of coating quality made by experience d operators based on general features of the impedance spectra and paramete rs such as breakpoint frequency f(b) and pore resistance R-po on the one ha nd and the classification results obtained from the ANNs on the other hand was obtained. Evaluation of the results of these analyses was made easier b y introduction of the coating quality index (CQI) which has Values between 0 and 1. Occasional discrepancies observed between classification results b ased on Phi-log f data vs. log /Z/-log f data occurred in the transition re gion between two types of classification, e.g. between 'intermediate' and ' poor'. These discrepancies have been explained based on the experimental da ta for R-po and f(b) and their time dependence. (C) 1999 Elsevier Science L td. All rights reserved.