liguge/Fault-diagnosis-for-small-samples-based-on-attention-mechanism
基于注意力机制的少量样本故障诊断 pytorch
Implements a BiGRU architecture with interpretable space-channel attention mechanisms and custom regularization (AMSGradP, 1D-Meta-ACON) designed for 1D vibration signal classification from bearing fault datasets. Key contributions include a novel 1D signal attention block, Global Average Pooling integration after BiGRU layers, and 1D-Grad-CAM++ for model interpretability in low-data regimes. Built on PyTorch with domain adaptation support via AdaBN for cross-machine generalization.
278 stars. No commits in the last 6 months.
Stars
278
Forks
13
Language
Python
License
—
Category
Last pushed
Jun 26, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/liguge/Fault-diagnosis-for-small-samples-based-on-attention-mechanism"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
petrobras/BibMon
Python package that provides predictive models for fault detection, soft sensing, and process...
hustcxl/Deep-learning-in-PHM
Deep learning in PHM，Deep learning in fault diagnosis，Deep learning in remaining useful life prediction
kokikwbt/predictive-maintenance
Datasets for Predictive Maintenance
biswajitsahoo1111/rul_codes_open
This repository contains code that implement common machine learning algorithms for remaining...
tvhahn/weibull-knowledge-informed-ml
Using knowledge-informed machine learning on the PRONOSTIA (FEMTO) and IMS bearing data sets....