On the Application of Cumulant-based Cyclostationary Processing on Bearings Diagnosis

A false indication of failure in rolling element bearings can be reached when cyclostationary processing technique is used for machine condition diagnosis. This problem is due to the fact that the estimated second-order cyclostationary parameters can be altered by first-order cyclostationary signals such as vibrations no related to those produced by defective bearings. The goal of this work is to solve this problem by applying a cumulant-based approach. Four algorithms were implemented. In order to quantify the effectiveness of the algorithm applications, a new function, named Interference Rate Function, is proposed. The appreciable interference immunity in the estimated cumulant-based cyclostationary parameters demonstrated the veracity of the hypothesis.