Heat Transfer Efficiency Prediction of Coal-Fired Power Plant Boiler Based on CEEMDAN-NAR Considering Ash Fouling

Ash fouling has been an important factor in reducing the heat transfer efficiency and safety of coal-fired power plant boilers. Scientific accurate prediction ash surfaces is basis formulating a reasonable soot blowing strategy to improve energy efficiency. This study presented comprehensive approach dynamic At first, cleanliness used reflect level surfaces. Then, model proposed predict deposits boilers by combining complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) nonlinear autoregressive neural networks (NARNN). To construct network model, minimum information criterion trial-and-error method are determine delay orders hidden layers. Finally, experimental object established on 300 MV economizer clearness dataset station, root mean square error absolute percentage smallest. In addition, results show that this multiscale more competitive than Elman model.