An improved method for restraining the end effect in empirical mode decomposition and its applications to the fault diagnosis of large rotating machinery

The purpose of this paper is to introduce an improved slope-based method (ISBM) to restrain the end effect in empirical mode decomposition (EMD). With this method, non-stationary, nonlinear time series can be decomposed efficiently and accurately into a set of intrinsic mode functions (IMFs) and a residual trend. Furthermore, due to its robust end effect restraint ability, the ISBM provides an attractive alternative to the traditional end condition methods. For the purpose of mechanical fault diagnosis, the IMFs derived from the improved EMD are then used to extract the features of faults and remove the interference from environmental noise and some irrelevant components. Industrial case studies on large rotating machinery show that IMF derived from improved EMD is relatively easy to understand and especially useful for analysis of non-stationary, nonlinear time series.