LAUSR

Shipborne antennas often undertake the tasks of guaranteeing ground-to-air communication. Rolling bearings, as key components of the shipborne antenna transmission system, improving its self-maintenance ability is an important link to guarantee the pointing accuracy of the entire antenna system. However, the lack of data, especially labeled data, typically hinders the wide application of intelligent fault diagnosis methods. To address this issue, a metalearning network is specially designed for intelligent fault identification of the bearings under the small samples prerequisite, which is named the affiliation network (AN). The AN consists of a random sampler, a feature extractor, an auxiliary classifier, and a discriminator. The former three are utilized to extract and concatenate the features from training and testing samples, while the latter trains an adaptive pseudodistance to evaluate the affiliation degree between concatenated features for identifying unknown data. Besides, a prior sufficient metatraining strategy is specially designed to realize metric-based knowledge transfer for acquiring the more generic AN in different application scenarios. The effectiveness of the proposed method is validated by three experimental cases. Results indicate that, compared with the state-of-the-art diagnostic models, the prior trained AN only utilized few samples to effectively identify failure categories of rolling bearings even with the complex operating conditions.