Condition Monitoring using Machine Learning: A Review of Theory, Applications, and Recent Advances

Surucu, Onur; Gadsden, S. Andrew; Yawney, John

doi:10.1016/j.eswa.2023.119738

Cited by 105 publications

(26 citation statements)

References 101 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, automated maintenance has gained a prominent role in the industrial environment [ 8 ]. Specifically, the use of informatics system for the automated detection of faults has been explored in many research fields.…”

Section: Introductionmentioning

confidence: 99%

Spectrogram Inversion for Reconstruction of Electric Currents at Industrial Frequencies: A Deep Learning Approach

Lalla,

Albini,

Di Barba

et al. 2024

Sensors

View full text Add to dashboard Cite

In this paper, we present a deep learning approach for identifying current intensity and frequency. The reconstruction is based on measurements of the magnetic field generated by the current flowing in a conductor. Magnetic field data are collected using a magnetic probe capable of generating a spectrogram, representing the spectrum of frequencies of the magnetic field over time. These spectrograms are saved as images characterized by color density proportional to the induction field value at a given frequency. The proposed deep learning approach utilizes a convolutional neural network (CNN) with the spectrogram image as input and the current or frequency value as output. One advantage of this approach is that current estimation is achieved contactless, using a simple magnetic field probe positioned close to the conductor.

show abstract

Section: Introductionmentioning

confidence: 99%

Spectrogram Inversion for Reconstruction of Electric Currents at Industrial Frequencies: A Deep Learning Approach

Lalla,

Albini,

Di Barba

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…SVR is effective in handling high-dimensional data, and is robust to outliers. All eight models described can be considered classic machine learning algorithms commonly applied to solve regression problems [ 28 , 29 , 30 ]. Advantageously, machine learning can handle extensive and complex datasets that would be difficult or impossible to analyze manually [ 31 ].…”

Section: Introductionmentioning

confidence: 99%

Unraveling the Complex Interactions: Machine Learning Approaches to Predict Bacterial Survival against ZnO and Lanthanum-Doped ZnO Nanoparticles

Navarro-López,

Perfecto-Avalos,

Zavala

et al. 2024

Antibiotics

View full text Add to dashboard Cite

The rise in antibiotic-resistant bacteria is a global health challenge. Due to their unique properties, metal oxide nanoparticles show promise in addressing this issue. However, optimizing these properties requires a deep understanding of complex interactions. This study incorporated data-driven machine learning to predict bacterial survival against lanthanum-doped ZnO nanoparticles. The effect of incorporation of lanthanum ions on ZnO was analyzed. Even with high lanthanum concentration, no significant variations in structural, morphological, and optical properties were observed. The antibacterial activity of La-doped ZnO nanoparticles against Gram-positive and Gram-negative bacteria was qualitatively and quantitatively evaluated. Nanoparticles induce 60%, 95%, and 55% bacterial death against Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus, respectively. Algorithms such as Multilayer Perceptron, K-Nearest Neighbors, Gradient Boosting, and Extremely Random Trees were used to predict the bacterial survival percentage. Extremely Random Trees performed the best among these models with 95.08% accuracy. A feature relevance analysis extracted the most significant attributes to predict the bacterial survival percentage. Lanthanum content and particle size were irrelevant, despite what can be assumed. This approach offers a promising avenue for developing effective and tailored strategies to reduce the time and cost of developing antimicrobial nanoparticles.

show abstract

“…The advances in the field of machine learning continuously motivate novel methods for data-driven process monitoring. [8][9][10] With the help of kernel function mapping, the aforementioned typical linear algorithms such as PCA, ICA, and PLS can be easily extended to model the nonlinear characteristics of a given dataset. [11][12][13] Because of the salient nonlinear feature extraction capability, artificial neural network models such as auto-encoder, 14,15 convolutional neural network, 16 and longshort-term memory neural network, 17 have also demonstrated their powerful ability to fault detection.…”

Section: Introductionmentioning

confidence: 99%

“…Different from first‐principle‐model‐based methods that generate residuals to indicate a fault, 7 data‐driven process monitoring approaches mainly focus on modeling normal variation in a dataset given from the NOC. The advances in the field of machine learning continuously motivate novel methods for data‐driven process monitoring 8–10 . With the help of kernel function mapping, the aforementioned typical linear algorithms such as PCA, ICA, and PLS can be easily extended to model the nonlinear characteristics of a given dataset 11–13 .…”

Section: Introductionmentioning

confidence: 99%

Distributed statistical process monitoring based on block‐wise residual generator

Tong,

Zhou,

Qian

et al. 2023

Journal of Chemometrics

View full text Add to dashboard Cite

The increasing scale of modern chemical plants keeps popularizing investigation as well as application of distributed process monitoring approaches. With a goal of directly quantifying the normal relations between different blocks divided from the whole process, a novel multi‐block modeling strategy called block‐wise residual generator is proposed, which trains a residual generator for each block through using the partial least squares algorithm with single one output, so that the relation between the corresponding block and the others is quantified as a regression model in a block‐wise manner. The deviations caused by the abnormal samples to the normal relations quantified for different blocks could thus be efficiently captured by the residuals generated from the block regression models, which then provide sensitive information for fault detection and contribution‐based fault diagnosis. Moreover, the proposed method is applicable for both disjoint and overlapped block divisions, and the direct consideration of individually quantifying relations between different blocks can always guarantee its salient monitoring performance, as validated through comparisons with classical distributed process monitoring methods.

show abstract

Condition Monitoring using Machine Learning: A Review of Theory, Applications, and Recent Advances

Cited by 105 publications

References 101 publications

Spectrogram Inversion for Reconstruction of Electric Currents at Industrial Frequencies: A Deep Learning Approach

Spectrogram Inversion for Reconstruction of Electric Currents at Industrial Frequencies: A Deep Learning Approach

Unraveling the Complex Interactions: Machine Learning Approaches to Predict Bacterial Survival against ZnO and Lanthanum-Doped ZnO Nanoparticles

Distributed statistical process monitoring based on block‐wise residual generator

Contact Info

Product

Resources

About