An Adaptive VMD Method Based on Improved GOA to Extract Early Fault Feature of Rolling Bearings
International Journal of Innovative Computing, Information and Control
In order to identify the early fault of bearing, an early feature extraction method based on adaptive variational mode decomposition (VMD) is proposed. The method not only improves the local optimal problem of grasshopper optimization algorithm (GOA) but also can adaptively determine the mode number and penalty parameter of VMD. Firstly, the convex-concave decreasing strategy is introduced to adjust the decreasing coefficient of GOA. Then, energy entropy mutual information (EEMI) index is
... uced to consider the energy distribution of modes and the dependence between modes and the original signal. Secondly, the optimal parameters of VMD matching with the input signal are obtained by taking the maximum EEMI as the objective function. Finally, the bearing signal is decomposed by the optimized VMD and the sensitive mode with maximum kurtosis is determined, and the fault feature contained in the sensitive mode can be extracted by envelope demodulation. The optimization experiments of 23 sets of benchmark functions show that the convex-concave strategy enhances the balance between exploration and exploitation, and the global and local search ability and stability of the GOA are improved. The experiments on simulation signal and bearing signal show that IGOA-VMD has better decomposition performance than VMD with fixed parameters and GOA-VMD. Therefore, this method provides a new idea and solution for fault feature extraction of bearing and other key components.