A5066043427- Machine Fault Diagnosis Techniques
- Engineering Diagnostics and Reliability
- Gear and Bearing Dynamics Analysis
- Geophysical Methods and Applications
- Natural Language Processing Techniques
- Non-Destructive Testing Techniques
- Anomaly Detection Techniques and Applications
- Fault Detection and Control Systems
- Ultrasonics and Acoustic Wave Propagation
- Image Retrieval and Classification Techniques
- Handwritten Text Recognition Techniques
Qilu University of Technology
2022-2024
Shandong Academy of Sciences
2022-2024
In the event of mechanical equipment failure, fault may not belong to any known category, and existing deep learning methods often misclassify such faults into a class, leading erroneous diagnosis. order address challenge identifying new types in diagnosis, this paper proposes novelty detection diagnosis method for bearing based on hybrid autoencoder network. Firstly, network with one input two outputs was constructed. The original data were then fed obtain its low-dimensional representation...
In order to realize rapid fault detection and early warning, a method based on normal operation data is proposed. Firstly, the model constructed improved deep neural network of auto-encoder. Secondly, unsupervised pretraining supervised fine-tuning are finished through in state solve contradiction between small sample large training required by model. The adaptive threshold reconstruction error used as evaluation index reduce influence environmental factors. Experimental results show that...
Deep learning uses mechanical time-frequency signals to train deep neural networks, which realizes automatic feature extraction and intelligent diagnosis of fault features gets rid the dependence on a large number signal processing technology experience. Aiming at problem misclassification similar samples, algorithm based adaptive hierarchical clustering subset (AHC-SFD) is proposed extract applied gearbox diagnosis. Firstly, used analyze characteristics different data, then data set...
In this study, a novel neural network model (mCNN-LFLBs) combining multi-scale convolutional networks(mCNN) with local feature extraction blocks (LFLBs) is proposed for natural gas pipeline leakage systems based on distributed feedback fiber laser vibration sensors (DFB-FL). To address the problem of loss in traditional signal image transfer process, one-dimensional original directly used as input model, effectively maintaining time-domain correlation leaked signal. The preserved. that...