Application of a continuous oil product quality analysis using neural networks

In this article the problem of improving the accuracy of virtual analyzers for petroleum products quality determination was researched. Virtual analyzers increase the quality of products data generation on efficiency to management problem solving and optimization at the pace of the process. The goal was to develop the virtual analyzers of sulfuric acid alkylation unit main product based on neural networks to evaluate their accuracy and to compare with the results of virtual models of analyzers based on regression analysis.