Evidence of Chaotic Behavior in Noise From Industrial Process

Usually, in dealing with the dynamical behavior of noise from an industrial process (NFIP), random or even white noise is assumed to describe it. However, since white noise is just an ideal model in mathematics and can hardly exist in an industrial process, this assumption seriously hinders the successful application of classical control theory in practice. For this reason, nonlinear time series analysis is used to analyze NFIP in an attempt to capture the signature of chaos. The chaotic behavior is apparent in NFIP due to the presence of finite and non-integer correlation dimension, and the minimum number of macroscopic variables required to model NFIP dynamics is obtained. These results indicate that NFIP is deterministic and high-dimensional chaos but not random or white noise and, in the future, might lead the way towards the direct solution of prediction and modeling of NFIP and a series of changes to the traditional algorithms in the control field may appear