A5060062928- Machine Fault Diagnosis Techniques
- Gear and Bearing Dynamics Analysis
- Industrial Vision Systems and Defect Detection
- Fault Detection and Control Systems
- Anomaly Detection Techniques and Applications
- Integrated Circuits and Semiconductor Failure Analysis
- Engineering Diagnostics and Reliability
- Structural Health Monitoring Techniques
- Image and Signal Denoising Methods
- Advanced Algorithms and Applications
Wuhan University of Science and Technology
2022-2024
Deakin University
2023-2024
Existing intelligent fault diagnosis approaches demand substantial data for training diagnostic models. However, factors such as the inherent characteristics of bearings, operating conditions, and privacy security make collecting comprehensive fault-bearing very difficult. Although generating synthetic through generative adversarial networks (GANs) is feasible, generation GANs a time-consuming process. To address these challenges, framework based on GAN deep transfer learning (DTL) proposed,...
Recently, deep learning models have been widely studied and applied in fault diagnosis. However, two common drawbacks are: 1) they usually require a large amount of storage resources, making it difficult to run them on embedded devices 2) there is no access sufficient reliable training data train comprehensive diagnosis model. In this study, fusion approach proposed based knowledge distillation generative adversarial network (GAN). This named small-sample dense teacher assistant (SS-DTAKD),...
Abstract Generative adversarial networks (GANs) have shown promise in the field of small sample fault diagnosis. However, it is worth noting that generating synthetic data using GANs time-consuming, and cannot fully replace real data. To expedite GAN-based diagnostics process, this paper proposes a hybrid lightweight method for compressing GAN parameters. First, three modules are constructed: teacher generator, discriminator, student based on knowledge distillation (KD-GAN) approach. The...
Non-linear behavior is widespread in many kinds of signals from nature and engineering fields. Although the high energy-concentration level various advanced time-frequency (TF) analysis (TFA) techniques currently developed ensure a fine representation non-linear time-varying component (TVC) signal, it far sufficient to solely consider single aspect level, because actual signal composition always more complicated, especially for some thorny difficulties such as strong noise interference early...
Abstract Compared with signals collected by the single sensor, multivariate contain more information to reflect state of mechanical equipment, which has a positive effect on fault diagnosis. However, different acquisition channels and various operating conditions interfere extraction features rotating machinery. To solve this problem, taking rolling bearings as an example in paper, novel method is adopted alleviate these interferences combined improved extreme learning machine (ELM) achieve...
Bearing fault detection and classification under a diagnostics model with fewer parameters has been challenging problem. A common solution is knowledge distillation (KD) using teacher–student models. Through the process, student can acquire from teacher to enhance performance without introducing extra parameters. However, when powerful model, not always ideal. This because more generate specific strategies, which may result in poorer performance. To this end, multiassistant KD (MAKD) method...