DTST: A Dual-Aspect Time Series Transformer Model for Fault Diagnosis of Space Power System

Xu, Zhiqiang; Du, Mingyang; Zhang, Yujie; Miao, Qiang

doi:10.1109/tim.2024.3396856

Cited by 1 publication

(1 citation statement)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shi et al [17] combined convolution neural network and long short-term memory network (LSTM) to conduct fault diagnosis. Moreover, some advanced models such as temporal convolution network (TCN) [18] and transformer [19] have excellent capabilities in temporal feature extraction and global association learning [20], which have been applied in fault diagnosis. Speed and torque signals are also monitored in practice, but the aforementioned studies do not consider potential fault features contained within these signals, and these potential features can effectively enhance the representation capability of induction motor faults.…”

Section: Introductionmentioning

confidence: 99%

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*

Niu,

Wang,

Miao

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Induction motors play a vital role in the cooling water supply system of hydropower facilities. However, it is unrealistic to collect suffi-cient fault samples in hydropower station. The scarcity of labeled samples poses a challenge in developing a powerful diagnostic model with high classification accuracy. To address this challenge, this paper proposes a multi-channel data fusion strategy based on Trans-former for feature enhancement. Initially, original signals are transferred into non-overlapping single-channel data patches to preserve correlation features across different channels. Then, temporal and spatial attention modules are applied to process the data patches, which can learn and fuse temporal and spatial information, respectively. Subsequently, the data patches are embedded to retain posi-tion information and represent fault-related features through class embedding, which are further processed by a Transformer encoder with self-attention mechanisms. Finally, the classification task is achieved by utilizing a multilayer perceptron layer connected to the class embedding. In facing up with limited training samples, the proposed method could learn robust features that are beneficial to improve the fault diagnosis ability of induction motor. The proposed method is compared with three basic models and two improved methods, demonstrating the superiority of the proposed method in accuracy and feature clustering performance with limited sample conditions. Additionally, the ablation experiment demonstrates the necessity of each module in the proposed method.

show abstract

Section: Introductionmentioning

confidence: 99%

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*

Niu,

Wang,

Miao

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

DTST: A Dual-Aspect Time Series Transformer Model for Fault Diagnosis of Space Power System

Cited by 1 publication

References 32 publications

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*

Contact Info

Product

Resources

About

DTST: A Dual-Aspect Time Series Transformer Model for Fault Diagnosis of Space Power System

Cited by 1 publication

References 32 publications

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions*

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions*

Contact Info

Product

Resources

About

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*

Fault diagnosis of induction motor in the cooling water supply system using a multi-channel data fusion transformer with limited sample conditions^*