Fault diagnosis of bearings using machine learning algorithm

이준혁,; 유승열,; 신성철,; 강동훈,; 이순섭,; 이재철,

doi:10.5916/jkosme.2019.43.6.455

Cited by 11 publications

(9 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It attempts to use a reasonable number of parameters to obtain better graph classification performance. Inspired by the work of [23], we adopted a graph pooling method based on self-attention, called SAG-Pool, to extract crucial features from EEG data. The detailed steps are shown as follows:…”

Section: Gannsmentioning

confidence: 99%

DAGAM: a domain adversarial graph attention model for subject-independent EEG-based emotion recognition

Xu¹,

Dang²,

Wang³

et al. 2023

J. Neural Eng.

View full text Add to dashboard Cite

Objective. Due to individual differences in EEG signals, the learning model built by the subject-dependent technique from one person's data would be inaccurate when applied to another person for emotion recognition. Thus, the subject-dependent approach for emotion recognition may result in poor generalization performance when compared to the subject-independent approach. Existing studies, However, have attempted but not fully utilized EEG's topology, nor have they solved the problem caused by the difference in data distribution between the source and target domains. Approach. To eliminate individual differences in EEG signals, this paper proposes the domain adversarial graph attention model (DAGAM), a novel EEG-based emotion recognition model. The basic idea is to generate a graph using biological topology to model multichannel EEG signals. Graph theory can topologically describe and analyze EEG channel relationships and mutual dependencies. Then, unlike other graph convolutional networks, self-attention pooling is used to benefit the extraction of salient EEG features from the graph, effectively improving performance. Finally, following graph pooling, the domain adversarial based on the graph is used to identify and handle EEG variation across subjects, achieving good generalizability efficiently. Main Results. We conduct extensive evaluations on two benchmark datasets (SEED and SEED IV) and obtain cutting-edge results in subject-independent emotion recognition. Our model boosts the SEED accuracy to 92.59% (4.06% improvement) with the lowest standard deviation of 3.21% (2.46% decrements) and SEED IV accuracy to 80.74% (6.90% improvement) with the lowest standard deviation of 4.14% (3.88% decrements) respectively. The computational complexity is drastically reduced in comparison to similar efforts (33 times lower). Significance. We have developed a model that significantly reduces the computation time while maintaining accuracy, making EEG-based emotion decoding more practical and generalizable.

show abstract

Section: Gannsmentioning

confidence: 99%

DAGAM: a domain adversarial graph attention model for subject-independent EEG-based emotion recognition

Xu¹,

Dang²,

Wang³

et al. 2023

J. Neural Eng.

View full text Add to dashboard Cite

show abstract

“…Graph coarsening and pooling layers are important for GNN models to avoid overfitting by reducing the number of parameters [45]. Self-attention graph pooling (SAGPooling) [45,46] aims to aggregate the information of a graph's nodes and reduce the graph's size for further processing. SAGPooling is a type of graph pooling which is based on self-attention mechanisms.…”

Section: Gnn Backgroundmentioning

confidence: 99%

“…Where idx is the index of the new set of selected nodes and top k is the function that selects the top k nodes based on their attention scores as shown in figure 2(c). The remaining nodes' feature matrix and adjacency matrix are updated with X (l+1) = X (l) ⊙ Z idx and Â(l+1) = Â(l) idx,idx , where ⊙ is the elementwise product [45].…”

Section: Gnn Backgroundmentioning

confidence: 99%

“…Each GCN layer performs graph convolution and produces the output X ′ . After each graph convolution layer, we employ batch normalization and the SAGPooling strategy [45] to learn the top k most important nodes and drop the rest of the nodes. In the final layer of our GNN block, We used global mean pooling which averages the features across the remaining nodes and flattens the graph feature as a latent vector z.…”

Section: Temporal/spatial Feature Learningmentioning

confidence: 99%

See 1 more Smart Citation

EEG decoding for datasets with heterogenous electrode configurations using transfer learning graph neural networks

Han,

Wei,

Faisal

2023

J. Neural Eng.

View full text Add to dashboard Cite

Objective. Brain-machine interfacing (BMI) has greatly benefited from adopting machine learning methods for feature learning that require extensive data for training, which are often unavailable from a single dataset. Yet, it is difficult to combine data across labs or even data within the same lab collected over the years due to the variation in recording equipment and electrode layouts resulting in shifts in data distribution, changes in data dimensionality, and altered identity of data dimensions. Our objective is to overcome this limitation and learn from many different and diverse datasets across labs with different experimental protocols. Approach. To tackle the domain adaptation problem, we developed a novel machine learning framework combining graph neural networks (GNNs) and transfer learning methodologies for non-invasive motor imagery (MI) EEG decoding, as an example of BMI. Empirically, we focus on the challenges of learning from EEG data with different electrode layouts and varying numbers of electrodes. We utilize three MI EEG databases collected using very different numbers of EEG sensors (from 22 channels to 64) and layouts (from custom layouts to 10–20). Main results. Our model achieved the highest accuracy with lower standard deviations on the testing datasets. This indicates that the GNN-based transfer learning framework can effectively aggregate knowledge from multiple datasets with different electrode layouts, leading to improved generalization in subject-independent MI EEG classification. Significance. The findings of this study have important implications for brain-computer-interface research, as they highlight a promising method for overcoming the limitations posed by non-unified experimental setups. By enabling the integration of diverse datasets with varying electrode layouts, our proposed approach can help advance the development and application of BMI technologies.

show abstract

“…Where α ij is expressed as: The update function U is expressed as: The pooling mechanism of TopK [20][21][22] is a process of continuously discarding nodes according to the characteristic data of different scales on the It places the pooling scope on the full graph node. By setting the pooling rate k, k ∈ (0, 1), then learning to obtain a value z that can characterize the node importance and sorting it.…”

Section: Model Of a Troubleshooting Algorithm Based On Stack Diagram ...mentioning

confidence: 99%

Multisource electrohydraulic servo valve fault status diagnostic algorithm based on a message propagation mechanism

Gao

sun

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Fault identification of electrohydraulic servo valves is crucial to maintain the reliability and safety of high-precision electrohydraulic servo systems. Because the nonlinear characteristics and fault characteristics of electrohydraulic servo systems under noise conditions are implicit, it is difficult to obtain a large number of fault data of electrohydraulic servo valves. Therefore, an electrohydraulic servo valve fault diagnosis model based on characteristic distillation is proposed in this paper. First, the original fault data model is obtained based on an electrohydraulic servo valve fault test platform, the data are standardized, and the data of more than one cycle are extracted using a combination of down sampling and a sliding window for data enhancement. Second, a neural network fault diagnosis algorithm based on stack graph convolution is proposed, which is suitable for detecting different types of states (normal state, wear state, stuck state and coil short-circuit state) of electrohydraulic servo valves. The accuracy of the test set fluctuates between 0.7 and 1.0. Then, because there is a certain relationship between the characteristic smoothing phenomenon of a stack graph convolution model and the number of layers, a multilayer stack graph convolution model is bound to have problems such as model degradation. Therefore, a residual model is introduced into the stack model to improve the convergence speed of the model during the optimization process. The results show that the average accuracy of this method is 100%.

show abstract

Fault diagnosis of bearings using machine learning algorithm

Cited by 11 publications

References 0 publications

DAGAM: a domain adversarial graph attention model for subject-independent EEG-based emotion recognition

DAGAM: a domain adversarial graph attention model for subject-independent EEG-based emotion recognition

EEG decoding for datasets with heterogenous electrode configurations using transfer learning graph neural networks

Multisource electrohydraulic servo valve fault status diagnostic algorithm based on a message propagation mechanism

Contact Info

Product

Resources

About