Recurrent neural networks for long-term prediction in machine condition monitoring

Malhi, A.; Gao, Robert X.

doi:10.1109/imtc.2004.1351492

Cited by 8 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Seker et al (2003) have studied the ability of Elman's recurrent neural network to diagnose faults and conditions of a nuclear power plant and rotating machine structures. Malhi and Gao (2004) approached recurrent neural networks to predict the condition by monitoring of machines. Using the modal properties, a neural network method was proposed for damage identification in a bridge structure by Lee et al (2005) by considering errors occurring in baseline finite element models.…”

Section: Introductionmentioning

confidence: 99%

Structural damage detection in moving load problem using JRNNs based method

Jena¹,

Parhi²

2019

Journal of Theoretical and Applied Mechanics

View full text Add to dashboard Cite

Damage detection in a structure using the vibration signature is a quiet smart method for condition monitoring of the structure. In this problem, the Recurrent Neural Networks (RNNs) based method has been implemented for damage detection in the moving load problem as an inverse method. A multi-cracked simply supported beam under a traversing load has been considered for the present problem. The localization and severities of the supervised cracks on the structure are determined using the adapted Jordan's Recurrent Neural Networks (JRNNs) approach. The mechanism of Levenberg-Marquardt's back propagation algorithm has been implemented to train the networks. To check the adoptability of the proposed JRNNs method, numerical analyses along with laboratory test verifications have been conducted and found to be well emerged.

show abstract

Section: Introductionmentioning

confidence: 99%

Structural damage detection in moving load problem using JRNNs based method

Jena¹,

Parhi²

2019

Journal of Theoretical and Applied Mechanics

View full text Add to dashboard Cite

show abstract

“…Within the deep learning architecture, the recurrent neural networks (RNNs) can mainly handle temporal data analysis that prompted researchers to applied it for the industrial PHM process. Some researchers [12], [13], [14] proposed an RNNbased methods toward the prognostic issue. However, RNNs had the vanishing gradient or exploding problem arising in long sequence input, which cannot keep the previous information, except only the latest one.…”

Section: A Data-driven Methods For Rul Estimationmentioning

confidence: 99%

RUL Prediction Using a Fusion of Attention-Based Convolutional Variational AutoEncoder and Ensemble Learning Classifier

et al. 2023

View full text Add to dashboard Cite

Predicting the remaining useful life (RUL) is a critical step before the decision-making process and developing maintenance strategies. As a result, it is frequently impacted by uncertainty in a practical context and may cause issues. This paper proposes a new hybrid deep architecture that predicts when an in-service machine will fail to overcome the latter problem, allowing for an improved data analysis and dimensionality reduction capability providing better spatial distributions of features and increasing interpretability. A deep Convolutional Variational AutoEncoder with an Attention mechanism (ACVAE) has been developed and tested using the aero-engine C-MAPSS dataset. We defined two adapted threshold settings (α1, α2) by analysing the spatial distribution and minimizing the overlapping area between the degradation classes. To reduce the conflict zone, we used the soft voting classifier. The performance of our visual explainable deep learning model has reached a higher level of accuracy compared with previous existing models.

show abstract