Nonlinear Analysis of Experimental Noisy Time Series in Fluidized Bed Systems

The paper describes the application of some numerical techniques to analyze and to characterize the observed dynamical behaviour of fluidized bed systems. The preliminary results showed clearly that the dynamics of the considered process can be nonrecurrent and governed by chaotic deterministic rules rather than by stochastic ones. This significant conclusion was allowed by the application of a proper filtering procedure which was able to reduce the unwanted influence of significant broadband noise detected in the sampled experimental data.