Attributional Consistency-Driven Lightweight Incremental Learning Framework for Machine Fault Diagnosis in Industrial IoT
- 9 June 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. PP (16) , 1
Abstract
In industrial IOT environments, intelligent systems must continually learn from dynamic data streams to maintain diagnostic accuracy for emerging fault types. However, mainstream replay-based incremental learning methods typically depend on knowledge distillation with cumbersome teacher models to preserve past knowledge, resulting in excessive memory overheads that preclude their deployment on resource-constrained edge devices. To overcome this challenge, we propose ME-ACR, a memory-efficient framework driven by attributional consistency reinforcement. ME-ACR first employs cosine-normalized classifier to mitigate the inherent recency bias and establish a stable feature space. Then, two synergistic self-supervised paradigms are designed, which anchors old knowledge without a teacher model. Attributional consistency reinforcement paradigm enforces internal cohesion by promoting consistency among the attribution patterns of correctly classified exemplars. Prototype-distance-based attribution disentanglement paradigm ensures external separation by aligning the topological structure of attribution space with the global geometry of feature space. By preserving the structure of both intra-class and inter-class attributional knowledge, ME-ACR effectively prevents catastrophic forgetting. Extensive experimental results on multiple datasets demonstrate the superiority of the proposed ME-ACR in achieving robust and memory-efficient fault diagnosis under incremental learning scenarios.Keywords
Funding Information
- Scientific Research Foundation for High-level Talents of Anhui University of Science and Technology (2025yjrc0031)
- National Natural Science Foundation of China (52475116)
- National Natural Science Foundation of China (62476006)
- Jiangsu Qinglan Project
- Fundamental Research Funds of the AUST (2024JBZD0007)
This publication has 25 references indexed in Scilit:
- Continual learning for unknown domain fault diagnosis in rotating machinery via Diffusion-Integrated Dynamic Mixture ExpertsEngineering Applications of Artificial Intelligence, 2025
- HSE: A plug-and-play module for unified fault diagnosis foundation modelsInformation Fusion, 2025
- Compact-sparse prototype calibration network for few-shot continual fault diagnosis of rotating machineryEngineering Applications of Artificial Intelligence, 2025
- Dynamic branch layer fusion: A new continual learning method for rotating machinery fault diagnosisKnowledge-Based Systems, 2025
- A new lifelong learning method based on dual distillation for bearing diagnosis with incremental fault typesAdvanced Engineering Informatics, 2025
- A relationship-aware calibrated prototypical network for fault incremental diagnosis of electric motors without reserved samplesReliability Engineering & System Safety, 2024
- A multi-sensor fused incremental broad learning with D-S theory for online fault diagnosis of rotating machineryAdvanced Engineering Informatics, 2024
- Self-driven continual learning for class-added motor fault diagnosis based on unseen fault detector and propensity distillationEngineering Applications of Artificial Intelligence, 2023
- Adaptive incremental diagnosis model for intelligent fault diagnosis with dynamic weight correctionReliability Engineering & System Safety, 2023
- Continual learning fault diagnosis: A dual-branch adaptive aggregation residual network for fault diagnosis with machine incrementsChinese Journal of Aeronautics, 2022