A Few-Shot Learning Method Incorporating Graph Sample Augmentation for UAV Fault Detection With Signal Loss

Abstract

The shortage of labeled historical data, particularly the reduction of partial sensor signals in flight logs, has diminished the accuracy of UAV fault detection methods during long-term flights. The limited prior spatial information derived from scarce and incomplete historical data causes overfitting in detection models, particularly when addressing large-scale and heterogeneous online data. This article proposes a self-supervised prototypical network (SSPN) with a graph sample augmentation method (GSAM) to leverage a small amount of available training samples and enhance the generalization performance of the fault detectors. Missing sensor signals are reconstructed by exploiting the remaining sensor signals in the historical data to create complete monitoring samples. Subsequently, a subset of sensors is randomly removed from these complete samples, and additional samples are resampled from them to augment the training dataset. The augmented training samples are grouped and aggregated into multiple prototypes based on their categories. Online data are sequentially matched to the prototypes corresponding to various fault types and identified based on their similarity. For unlabeled unknown faults, a metatrained detector is designed to quickly learn and classify anomalies by utilizing prior knowledge from related metatasks. The experimental results, based on datasets from three UAVs, demonstrate the effectiveness of the proposed method.