Intelligent Fault Identification in Flight Vehicles Using Bayes Classifier with Dynamic Feature Optimization and Adaptive Model Selection

To address reliable and efficient fault identification in flight vehicles, this paper proposes a systematic framework integrating dynamic feature optimization with an adaptive Bayesian classifier. It enhances fault diagnosis performance via a multi-domain feature engineering pipeline for key feature extraction and an automated model selection process. Considering the dynamic and complex nature of flight vehicle data, the method effectively captures feature differences between fault modes, strengthening multi-type fault identification. Case study results show the optimized model significantly outperforms baseline classifiers with high fault identification accuracy.