Composite fault feature extraction of automotive transmission shaft bearings of optimizing RVME parameters based on SWO

To deal with the issue of composite fault diagnosis of automotive transmission shaft bearings under strong background noise interference, a feature extraction method of optimizing recursive variational mode extraction parameters based on spider wasp optimizer (SWO) is proposed. SWO is applied to determine the optimal values of mode extraction parameters. Effective weight kurtosis criterion is employed to reduce the background noise interference. The introduction of a two-dimensional Chebyshev-Logistic-Infinite collapse map increases the diversity and randomness of the population, shortens running time, and increases fault feature coefficient. The population is divided into multiple subgroups based on fast nondominated sorting, and an adaptive grouping strategy is adopted to accelerate the convergence speed of SWO. The results demonstrate the approach can suppress the interference noise with effect, and gain a good balance between search speed and feature extraction accuracy.