10 / 2025-03-13 20:55:38

A Fault Diagnosis Method for Multi-UAVs Based on Belief Rule Base with Wavelet Transformation

Belief rule base,fault diagnosis,multi-UAV system,wavelet transformation,gradient descent optimization

Fault diagnosis is a crucial problem for multiunmanned aerial vehicles (UAVs) under complex conditions. A new fault diagnosis method based on belief rule base with wavelet transformation (BRB-WT) is proposed in this paper, where a gradient descent optimization method is constructed to optimize hyperparameters of the proposed BRB-WT. Advantages of the proposed method are twofold: First, the proposed BRB-WT method could effectively capture characteristics, such as high-frequency components and rapid signal changes, from time-domain signals of UAVs. Second, the proposed gradient descent optimization method is more interpretable compared with some existing works of BRB with P-CMA-ES optimization method. These enhancements significantly improve the overall performance of the proposed fault diagnosis method based on BRB-WT. Case studies and comparative analyses demonstrate that the proposed BRB-WT method outperforms existing fault diagnosis techniques in terms of diagnostic accuracy.