AN ANFIS BASED SYSTEM FOR PREDICTING WALL TEMPERATURE OF SUPERCRITICAL BOILERS

The ever increasing energy demand has necessitated the wide spread adoption of supercritical boilers. These boilers which operate at extremely high temperature and pressure have thrown up a number of challenges that have to be surmounted. One such challenge is to measure the wall temperature of tubes used in these supercritical boilers to assess efficiency. Conventional methods are either complicated or insufficient technically to predict the wall temperature of the metal tubes used in these boilers. In this work, an expert system has been formulated and presented for prediction of wall temperatures of boiler tubes. An Adaptive Neuro Fuzzy Interference System (ANFIS) has been modeled to predict the wall temperature. The ANFIS based prediction is modeled on inputs with internal tube diameter, pressure inside the boiler tubes, heat flux, mass flux and bulk fluid temperature. The results of the proposed model are validated against the results of experimentation and also been validated by calculating the Normalized Root Mean Square (NRMS) value between the experimental value and the temperature predicted by the proposed model. The results of prediction are found to be in close comparison to that of experimentation.