A novel extraction method for useful component of vibration signal combining variational mode decomposition and relative entropy

Abstract

In condition monitoring and prognostics health management, it is very important to extract the useful components of equipment state signals. In this paper, combining variational mode decomposition (VMD) and relative entropy (RE), a novel approach is proposed for extracting signal useful components. By using VMD, the original vibration signal can be adaptively decomposed, and its effective constituents can be acquired through the assessment of RE. The proposed method is further applied into some simulated and measured signals of a hydraulic axial piston pump. The effectiveness and feasibility of the proposed method are demonstrated through the numerical and tested vibration signals. The results show that the proposed method possesses laudable capability to extract the effective component of vibration signals for a hydraulic axial piston pump under normal state, slipper wear, and slipper luxation. The interference of background noise is effectively overcome. Furthermore, the expected useful signals are precisely reconstituted.

Water quality prediction in sea cucumber farming based on a GRU neural network optimized by an improved whale optimization algorithm

“…Relative entropy is the quantification of the degree of difference between two probabilities ( Zhu et al, 2021 ). Relative entropy can measure the difference and closeness of two probability distributions, and can be used as the loss function of some optimization algorithms.…”