Abstract:Considering the changing working conditions of rotating machinery in operation, it is often difficult to collect data accurately in some severe fault states, and the lack of data can lead to poor performance of deep learning-based fault diagnosis models. In the few-shot scenario, traditional meta-learning methods yield superior predictions by identifying ideal initialization parameters. When the update gradients of tasks are in different directions, meta-learning modifies the weights of samples in the current tasks to average them out, which may lead to negative transfer between tasks and poor generalization of biases. Based on the above problems, we propose a meta-learning network with anti-interference (AIML) for few-shot fault diagnosis, which is obtained by combining a dynamic fine-tuning technique to increase the gradient agreement of the tasks. AIML consists of two functions: the feature encoding network (FEN) and the base network. AIML uses the property of shared parameters about transfer learning to learn shared feature representations of different tasks, while FEN dynamically adjusts the loss weights of conflicting tasks due to meta-learning with two-level optimization techniques. First, in the internal loop, AIML updates the gradient of the base network while the network parameters of FEN are fixed. Instead of immediately computing the gradient of the optimal parameters for the new task, the FEN is then integrated in the outer loop to learn the meta-representation, and the weights obtained for the various tasks relate to the gradient for meta-optimization. Only during the outer loop stage of the meta-training is the FEN updated. Three publicly available datasets are used to assess the performance of AIML, and the results show that it is more effective at resolving problems involving few-shot fault diagnosis.

A Prototype Correction Network with Coordinate Attention for Few-Shot Fault Diagnosis Under Varying Working Conditions

An improved prototypical network with L2 prototype correction for few-shot cross-domain fault diagnosis

A Fine-Tuning Prototypical Network for Few-shot Cross-domain Fault Diagnosis

Adaptive Generic Prototype Network with Geodesic Distance for Cross-Domain Few-Shot Fault Diagnosis

Semi-Supervised Prototype Network with CBAM and Data Selector for Few-Shot Bearing Fault Diagnosis

Uncertainty-based Contrastive Prototype-Matching Network Towards Cross-Domain Fault Diagnosis with Small Data

Domain-Adaptive Prototype-Recalibrated Network with Transductive Learning Paradigm for Intelligent Fault Diagnosis under Various Limited Data Conditions

Residual Prototype Network-based Method for Few-Shot Fault Diagnosis of Bearing

A Prototypical Networks-based Multi-task Model for Few-shot Fault Diagnosis

Dual prototypical contrastive network: a novel self-supervised method for cross-domain few-shot fault diagnosis

Few-Shot Learning for Fault Diagnosis: Semi-Supervised Prototypical Network with Pseudo-Labels

Prototype Space Boundary Alignment Network for Bearing Continuous Fault Diagnosis under Class-Incremental Scenarios

A Fault Diagnosis Method Using Improved Prototypical Network and Weighting Similarity-Manhattan Distance with Insufficient Noisy Data

Reweighted Regularized Prototypical Network for Few-Shot Fault Diagnosis

Generic Meta-Transfer Learning Model with Special Neuronal Processing Parameters for Few-Shot Fault Bearing Diagnosis

A Meta-Learning Network with Anti-Interference for Few-Shot Fault Diagnosis

PCDC: Prototype-assisted dual-contrastive learning with depthwise separable convolutional neural network for few-shot fault diagnosis of permanent magnet synchronous motors under new operating conditions

Attention-based Deep Meta-Transfer Learning for Few-Shot Fine-Grained Fault Diagnosis

Prototype augmented network with metric-mixed under limited samples for mechanical intelligent fault recognition.

Few-Shot Gearbox Fault Diagnosis Based on Semi-Supervised Prototypical Network and Contrastive Learning

Adaptive Neighborhood-Perceived Contrastive Network for Early Stage Fault Diagnosis of Rolling Bearing with Limited Labeled Data