Parameter-optimized variational mode decomposition based on bat algorithm for the fault diagnosis of rolling bearings

International audienceThe accurate fault diagnosis of rolling bearings is of great significance to ensure the safety of rotating machinery. Compared with the traditional methods, the variational mode decomposition (VMD) has many advantages on overcoming mode mixing and the inference of noise. However, the parameters of VMD need to be determined empirically which has great effects on the decomposition result. To address the problem, this paper proposes a parameter-optimized VMD method based on an improved bat algorithm for fault diagnosis of rolling bearings. The optimized parameters include the number of decomposition modes and the data-fidelity constraint α. Envelope entropy is used as the optimization objective for VMD, and the two parameters of VMD are optimized by an improved bat algorithm. With optimal parameters, the optimal band-limited intrinsic mode function (BLIMF) of VMD is obtained, and then the fault feature of the rolling bearing can be extracted by the envelope spectrum of the optimal BLIMF. Two experiments show that the proposed method can be effectively applied to the fault diagnosis of rolling bearings, and it is superior to other existing methods, such as empirical mode decomposition (EMD) and ensemble empirical mode decomposition (EEMD)