Monitoring of Fluidized Beds Hydrodynamics Using Recurrence Quantification Analysis

A new method is presented for on-line monitoring of fluidized beds hydrodynamics using pressure fluctuations signal by recurrence quantification analysis. The experiments were carried out at different gas velocities and sand types. A 95% confidence interval was computed for determinism (Det) of signals obtained from reference state as well as other operating conditions named as unideal states. Det of unideal states was compared with Det of the reference state to reject the null hypothesis that all the signals have been generated from the reference state. It was shown that Det is sensitive to small change in particles size whereas it is not sensitive to minor superficial gas velocity variations, indicating its ability for hydrodynamic on-line monitoring. Furthermore, in this method it is no need for time series embedding, long-term data sampling and time-consuming numerical algorithms. (C) 2012 American Institute of Chemical Engineers AIChE J, 59: 399406, 2013