KLASSIC: Knowledge Pool-Assisted Forgetting-Resistant Representation Calibration for Few-Shot Class-Incremental Fault Diagnosis

Machinery in continuous operation may develop new faults over time, which not only jeopardize the safe operation of equipment but also come with scant available fault data, hampering the establishment of reliable diagnostic models. These challenges, termed few-shot class-incremental fault diagnosis (FS-CIFD), require the model to balance between retaining past diagnostic abilities and adapting to new faults despite scarce data. Existing incremental fault diagnosis methods suffer from catastrophic forgetting, primarily due to the model’s representation bias toward newly occurring faults. Moreover, the scarcity of new fault data induces overfitting, further aggravating catastrophic forgetting. To tackle FS-CIFD, we propose a novel knowledge pool-assisted forgetting-resistant representation calibration (KLASSIC) method. Unlike existing methods, KLASSIC integrates signal statistical knowledge to enhance fault features in scarce data, enabling the model to learn generalized representations. First, a knowledge pool is established to offer a shared knowledge space and output task-relevant knowledge, reactivating the model’s memory of previously learned faults. Second, we design a knowledge-prompted-guided adaptive calibration (KPGAC) module to correct the representation bias toward new fault types and mitigate catastrophic forgetting. Third, to alleviate the overfitting, pseudo-incremental tasks are constructed, facilitating the development of a generalizable feature space for forthcoming tasks. Extensive experiments with different case studies demonstrate that KLASSIC outperforms existing methods in handling FS-CIFD challenges.

• Publication year

  2025

• Venue

  IEEE Transactions on Instrumentation and Measurement

• Publication date

  Unknown publication date

• Fields of study

  Not labeled

• Identifiers
  DOI 10.1109/TIM.2025.3628426
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Optimized forgetting factor recursive least square method for equivalent circuit model parameter extraction of battery and ultracapacitor

  2025cited by this paper
• A self-improving fault diagnosis method for intershaft bearings with missing training samples

  2025cited by this paper
• A Globally Interpretable Convolutional Neural Network Combining Bearing Semantics for Bearing Fault Diagnosis

  2025cited by this paper
• Open-Set Fault Diagnosis Based on 1D-ResNet With Fusion of Cross-Class and Extreme Information for Out-of-Distribution Detection

  2025cited by this paper
• Explainable fault diagnosis method based on statistical features for gearboxes

  2025cited by this paper
• A relationship-aware calibrated prototypical network for fault incremental diagnosis of electric motors without reserved samples

  2024cited by this paper
• A New Incremental Learning for Bearing Fault Diagnosis Under Noisy Conditions Using Classification and Feature-Level Information

  2024cited by this paper
• Multiscale Deep Attention Q Network: A New Deep Reinforcement Learning Method for Imbalanced Fault Diagnosis in Gearboxes

  2024cited by this paper
• TFPred: Learning discriminative representations from unlabeled data for few-label rotating machinery fault diagnosis

  2024cited by this paper
• Semi-supervised class incremental broad network for continuous diagnosis of rotating machinery faults with limited labeled samples

  2024cited by this paper
• A review on physics-informed data-driven remaining useful life prediction: Challenges and opportunities

  2024cited by this paper
• A broad learning model guided by global and local receptive causal features for online incremental machinery fault diagnosis

  2024cited by this paper
• Domain adaptation method based on pseudo-label dual-constraint targeted decoupling network for cross-machine fault diagnosis

  2024cited by this paper
• A prognostic model for multi-stage degraded equipment under zero life label combining CDBN and Bayesian bidirectional GRU

  2024cited by this paper
• Reserving embedding space for new fault types: A new continual learning method for bearing fault diagnosis

  2024cited by this paper
• Intelligent fault diagnosis of bearings driven by double-level data fusion based on multichannel sample fusion and feature fusion under time-varying speed conditions

  2024cited by this paper
• Semi-supervised prototype network based on compact-uniform-sparse representation for rotating machinery few-shot class incremental fault diagnosis

  2024cited by this paper
• Heterogeneous Federated Domain Generalization Network With Common Representation Learning for Cross-Load Machinery Fault Diagnosis

  2024cited by this paper
• A novel incremental method for bearing fault diagnosis that continuously incorporates unknown fault types

  2024influential reference
• Few-Shot Bearing Fault Diagnosis Via Ensembling Transformer-Based Model With Mahalanobis Distance Metric Learning From Multiscale Features

  2024influential reference
• Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration

  2023cited by this paper
• Physics-Informed Residual Network (PIResNet) for rolling element bearing fault diagnostics

  2023cited by this paper
• A Lifelong Learning Method Based on Generative Feature Replay for Bearing Diagnosis With Incremental Fault Types

  2023cited by this paper
• Gradient-Regulated Meta-Prompt Learning for Generalizable Vision-Language Models

  2023cited by this paper
• Class-Incremental Learning: A Survey

  2023cited by this paper
• Knowledge-informed deep networks for robust fault diagnosis of rolling bearings

  2023cited by this paper
• Three types of incremental learning

  2022influential reference
• Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks

  2022influential reference
• DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning

  2022cited by this paper
• A feature extraction and machine learning framework for bearing fault diagnosis

  2022cited by this paper
• A Lifelong Learning Method for Gearbox Diagnosis With Incremental Fault Types

  2022cited by this paper
• A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning

  2022cited by this paper
• MetaFSCIL: A Meta-Learning Approach for Few-Shot Class Incremental Learning

  2022cited by this paper
• Few-Shot Class-Incremental Learning via Entropy-Regularized Data-Free Replay

  2022cited by this paper
• Continual learning fault diagnosis: a dual-branch adaptive aggregation residual network for fault diagnosis with machine increments

  2022cited by this paper
• Challenges and Opportunities of AI-Enabled Monitoring, Diagnosis & Prognosis: A Review

  2021cited by this paper
• Few-Shot Class-Incremental Learning via Relation Knowledge Distillation

  2021cited by this paper
• Few-Shot Incremental Learning with Continually Evolved Classifiers

  2021cited by this paper
• Deep Learning Algorithms for Rotating Machinery Intelligent Diagnosis: An Open Source Benchmark Study

  2020influential reference
• Few-Shot Class-Incremental Learning

  2020cited by this paper
• Attentive Weights Generation for Few Shot Learning via Information Maximization

  2020cited by this paper
• Random Path Selection for Incremental Learning

  2019cited by this paper
• Maintaining Discrimination and Fairness in Class Incremental Learning

  2019cited by this paper
• Incremental Learning Using Conditional Adversarial Networks

  2019cited by this paper
• A novel intelligent method for bearing fault diagnosis based on affinity propagation clustering and adaptive feature selection

  2017cited by this paper
• Prototypical Networks for Few-shot Learning

  2017cited by this paper
• Attention is All you Need

  2017cited by this paper
• Overcoming catastrophic forgetting in neural networks

  2016cited by this paper
• iCaRL: Incremental Classifier and Representation Learning

  2016cited by this paper
• Chaotic dynamics and convergence analysis of temporal difference algorithms with bang‐bang control

  2016cited by this paper
• Matching Networks for One Shot Learning

  2016cited by this paper
• Deep Residual Learning for Image Recognition

  2015cited by this paper
• New clustering algorithm-based fault diagnosis using compensation distance evaluation technique

  2008cited by this paper
• Fault diagnosis of rotating machinery based on multiple ANFIS combination with GAs

  2007cited by this paper

• Polarization phase-shifting multi-wavelength interferometric displacement measurement with nonlinear error compensation

  2026cites this paper