Transparent and explicable boiler fouling monitoring with fuzzy neural newtwork

Abstract

Fouling on boiler heating surfaces is one of the important factors that damage boiler’s economical performance and safety. With on-line monitoring of foiling states on boiler heating surfaces, it is possible to optimize sootblower system, to visualize fouling states, to improve performance, as well as to remedy the insufficiency of experiment research in boiler heating surface fouling process. New method based on Fuzzy Neural Network (FNN) is presented to monitor fouling states on boiler heating surfaces on-line. Compared with regular methods, since FNN’s reasoning process is transparent and comprehensible, it is possible to explain and comprehend reasoning process, which makes the FNN based system perform as an additional operation consulting system.