Infrared Aircraft Classification Method with Small Samples Based on Improved Relation Network

Lu, 金璐 Jin; Shijian, 刘士建 Liu; Xiao, 王霄 Wang; Fanming, 李范鸣 Li

doi:10.3788/aos202040.0811005

Cited by 2 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each specification bearing data contain ten states including the normal state, inner raceway fault, outer raceway fault, rolling ball fault and different degrees of damage. IR (7,14,21) indicates that the damage diameters of the inner raceway are 7 mil, 14 mil and 21 mil respectively. Similarly, OR (7,14,21) and B (7,14,21) indicate different degrees of fault in the outer raceway and the rolling ball, respectively.…”

Section: Experimental Data Setmentioning

confidence: 99%

“…IR (7,14,21) indicates that the damage diameters of the inner raceway are 7 mil, 14 mil and 21 mil respectively. Similarly, OR (7,14,21) and B (7,14,21) indicate different degrees of fault in the outer raceway and the rolling ball, respectively. The vibration signals in the data set were truncated to samples with a length of 2048 points.…”

Section: Experimental Data Setmentioning

confidence: 99%

“…Multiple states Different loads SKF6205 (Meta-train set) N 0 hp (1797 r min −1 ) IR (7,14,21) 1 hp (1772 r min −1 ) OR (7,14,21) 2 hp (1750 r min −1 ) B (7,14,21) 3 hp (1730 r min −1 ) SKF6203 (Meta-test set) N 0 hp (1797 r min −1 ) IR (7,14,21) 1 hp (1772 r min −1 ) OR (7,14,21) 2 hp (1750 r min −1 ) B (7,14,21) 3 hp (1730 r min −1 ) classification results. When the meta-test set data only work under a single load, the classification effect is better.…”

Section: Bearing Specsmentioning

confidence: 99%

“…This method is highly applicable in the industrial field. Current mainstream solutions for fewshot learning tasks include transfer learning and meta-learning [14]. Transfer learning applies knowledge learned in a certain domain to solve problems in different but related domain.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Few-shot rolling bearing fault classification method based on improved relation network

Kang¹,

Liang²,

Wang³

et al. 2022

Meas. Sci. Technol.

View full text Add to dashboard Cite

In practical applications, it is difficult to obtain enough fault samples to train the fault classification model of rolling bearings, and the specifications of bearings used in different mechanical equipment may be different. The diagnosis model trained on a certain specification of bearings may not be applicable to other specification of bearings. To solve the above problems, a few-shot rolling bearing fault classification method based on improved relation network is proposed. Firstly, Fourier transform is applied to vibration signals of different specifications of bearings. The data of different specifications of bearings are divided into meta-train set and meta-test set according to the meta-learning training strategy, and each set is further divided into support set and query set. Secondly, an improved relation network is built. The residual shrinkage module and the SELU activation function are introduced into the embedding module of the relation network. The improved embedding module is used to extract the sample features of the support set and query set, and then the features of the two are combined and input into the relation module to get the relation score. The query set samples are classified according to the score. Finally, the rolling bearing fault classification model is obtained after multiple episodes. The experimental results show that, compared with the partial transfer learning and meta-learning methods, the proposed method only needs a few or even a single sample to achieve the fault classification of different specifications of rolling bearings under different loads. In the case of one-shot, the average classification accuracy can reach 93.3%.

show abstract

Section: Experimental Data Setmentioning

confidence: 99%