Optimised VMD based on Improved Grey Wolf for Human Pulse Wave Characterisation

Abstract

To address the difficulty in determining the penalty factor α mode and number K for the Variational Mode Decomposition (VMD) human pulse wave signal, which makes it difficult to extract and analyse its features accurately, an Improved Grey Wolf Optimization (IGWO) method is proposed to optimise the VMD. In this paper, the optimal modal number K and penalty factor α in the VMD are determined by IGWO using Minimum Average Mutual Information (MAMI) as the fitness function. The proposed method is used to completely decompose the pulse wave into K Intrinsic Mode Functions (IMF), and the marginal spectral analysis of the decomposed modes is performed to calculate the spectral energy ratio. And the effectiveness of the proposed method is demonstrated experimentally.