Unsupervised Bayesian change-point detection approach for reliable prognostics and health management of complex mechanical systems

Reliability Engineering & System Safety, Journal Year: 2024, Volume and Issue: 245, P. 110037 - 110037

Published: Feb. 24, 2024

Language: Английский

---

MNHP-GAE: A Novel Manipulator Intelligent Health State Diagnosis Method in Highly Imbalanced Scenarios

IEEE Internet of Things Journal, Journal Year: 2024, Volume and Issue: 11(13), P. 24073 - 24082

Published: April 16, 2024

As a classical and crucial component in industrial systems, the manipulators are widely employed precision manufacturing scenarios because of their advantages high stiffness, large load support capability, precision. During service, it is inevitable that they encounter data imbalance due to occasional low-frequency failure behaviors. But order address these issues, majority approaches already use need assistance extra tools. Thus, novel intelligent health state diagnosis model, named multiple neighbor homogeneous property-embedded graph auto-encoder (MNHP-GAE), developed get around this restriction apply manipulators. Its core realize expansion enrichment feature space by mining effective complementary information from property samples without augmentation other technologies. Specifically, wavelet decomposition reconstruction dynamic time warping integrated promote quantification sample similarity enable construction samples. Following that, unique module with multi-head attention mechanism constructed extract nodes match for diagnostic tasks. Finally, through multi-case experimental validation scenario 3-PRR planar parallel manipulator platform, superior performances proposed MNHP-GAE model highly unbalanced fully demonstrated.

Language: Английский

---

A review: the application of generative adversarial network for mechanical fault diagnosis

Measurement Science and Technology, Journal Year: 2024, Volume and Issue: 35(6), P. 062002 - 062002

Published: March 19, 2024

Abstract Mechanical fault diagnosis is crucial for ensuring the normal operation of mechanical equipment. With rapid development deep learning technology, methods based on big data-driven provide a new perspective machinery. However, equipment operates in condition most time, resulting collected data being imbalanced, which affects performance diagnosis. As approach generating data, generative adversarial network (GAN) can effectively address issues limited and imbalanced practical engineering applications. This paper provides comprehensive review GAN Firstly, GAN-based diagnosis, basic theory various variants (GANs) are briefly introduced. Subsequently, GANs summarized categorized from labels models, corresponding applications outlined. Lastly, limitations current research, future challenges, trends selecting application discussed.

Language: Английский

---

Knowledge Distillation-Guided Cost-Sensitive Ensemble Learning Framework for Imbalanced Fault Diagnosis

IEEE Internet of Things Journal, Journal Year: 2024, Volume and Issue: 11(13), P. 23110 - 23122

Published: July 1, 2024

In industrial scenarios, mechanical faults are episodic and uncertain. Thus the monitoring data collected is usually extremely imbalanced, resulting in intelligent diagnostic models that suffer from majority-class dominance, minority-class overfitting, poor generalization performance. Therefore, a knowledge distillation-guided cost-sensitive ensemble learning framework proposed. It effectively combines to fully extract multiscale features, leverage critical multi-depth emphasize classifying most confusing classes. Specifically, multiple-scale feature extraction multi-order fusion first employed utilize fault information. Afterward, complementary at different depths of network embedded into novel process for better integration decisions. Then an improved distillation method achieves mutual transfer sublimation excellent while focusing on classes achieve effective representation various types faults. Finally, strategy applied further increase attention minority The experimental results complex imbalance including extreme imbalance, step continuous interclass intra-class all indicate proposed can state-of-the-art performance provide promising solution practical application methods.

Language: Английский

---

A rolling bearing fault diagnosis method based on interactive generative feature space oversampling-based autoencoder under imbalanced data

Structural Health Monitoring, Journal Year: 2024, Volume and Issue: unknown

Published: May 10, 2024

With the rapid development of railroads and yearly increase in scale operation, safe operation maintenance rail trains have become particularly important. Among them, deep learning-based bearing fault diagnosis methods attracted more attention train maintenance. However, usually operate normally. Collecting complete data for learning model training is often difficult. Such scenarios with a large difference between number normal affect performance models. Here, an interactive generative feature space oversampling-based autoencoder (IGFSO-AE) proposed to realize sample generation under imbalanced data. First, original vibration signal converted into semantically stable amplitude–frequency by fast Fourier transform input autoencoder; second, order hidden layer features randomly exchanged, strategy then, interpolation oversampling used interpolate samples where Euclidean distance large, decoder, generated are mixed form new set, which intelligent output results. Finally, method evaluated using publicly available dataset bogie-bearing simulation bench our lab. The experimental results show that IGFSO-AE can generate diverse incremental information exhibits robustness superiority different proportions

Language: Английский

---