[1]Semi-Supervised Temporal Meta-Learning Framework for Wind Turbine Bearing Fault Diagnosis Under Limited Annotation Data
问题背景
- the fault data are so scarce that it is time-consuming to acquire a well behaved deep learning model
- much unlabeled data cannot be adequately utilized to explore useful fault information without prior.
方法思路
（1）a temporal convolutional module is proposed to relieve overfitting due to the depth of the model, which can fully excavate temporal features along the depth of the network.
（2）A novel deep learning generalization framework SeMeF—is proposed to make full use of massive unlabeled data and limited annotation data.
[2] A Meta-Learning Method for Electric Machine Bearing Fault Diagnosis Under Varying Working Conditions With Limited Data
问题背景
perform FD with a limited training data
方法思路
（1）a four layer CNN is used for feature learning and a simple convolution structure makes the training more effcient
(2) The meta-training process primarily completes the knowledge accumulation of prior tasks.
[3] Semi-supervised adaptive anti-noise meta-learning for few-shot industrial gearbox fault diagnosis
问题背景
obtain sufficient labeled data for FD is challenging
方法思路
（1）a residual network with a Morlet Wavelet layer is used to extract signal features
（2）sample-level attention is defined to select unlabeled samples that are more similar to labeled sample prototypes
（3）The adaptive metric is used to obtain the relational distance functions of labeled samples and unlabeled samples