Permutation entropy: A nonlinear statistical measure for status characterization of rotary machines

This paper investigates the usage of permutation entropy for working status characterization of rotary machines. As a statistical measure, the permutation entropy describes complexity of a time series or signal measured on a physical system through phase space reconstruction, and takes into account non-linear behavior of the time series, as often seen in vibration signals of rotary machines. Thus it can be served as a viable tool for detecting dynamic changes of the machine working status. The effect of embedded dimension and time delay on calculation of the permutation entropy value has been studied, and the validity of the permutation entropy for detecting dynamic change of a physical system is studied through a well known non-linear system, the Logistic Map. Comparison with other complexity measures using a numerically formulated signal has also been investigated. Experimental results on bearing vibration analysis have then confirmed that the permutation entropy provides an effective measure for monitoring the working status of rolling bearings.