Temporal deep learning framework for multi condition fault classification of three-phase induction motors

Abstract

Induction Motors (IMs) are a very important group of electrical appliances used in industry, due to their strength, versatility, high reliability, and cost-effectiveness. However, due to various faults, there can be many issues, such as production downtime, energy losses, and increased maintenance costs. This paper employs Deep learning (DL) frameworks, such as Simple Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and Gated Recurrent Unit (GRU), for the detection and categorization of IM. The novelty of the proposed work lies in the unified multi-sensor, multi-class, and multi-model DL framework with comprehensive cross-architecture evaluation for robust IM fault diagnosis under eight different operating scenarios, including startup transients and phase-removal faults, load-based faulty and healthy conditions. The results show that the DL is effective in electrical fault identification, using the classification accuracies of RNN—85.14%, LSTM −90.19%, BiLSTM—92.24%, and GRU—94.19%. The proposed approach can be used to differentiate between healthy and faulty operating conditions, which aim to highlight the effectiveness of DL for intelligent condition monitoring in induction motor drive-based machines.

KT-CBHNET: a spatio-temporal deep learning framework for interpretable ECG signal classification for cardiovascular disease prediction
Krantee Singh, Vasudev Dehalwar
Scientific Reports9 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Full textCardiovascular medicineCrossref
Traffic surveillance video-based a unified road accident detection, classification, and vehicle tracking framework using an optimized spatio-temporal attention mechanism
Chokkakula Devi, Gowri S.
PLOS One15 Oct 2026Peer-reviewedCC BY 4.0Full textEpidemiology & public healthPLOS
STCP-Net: A few-shot learning approach based on spectral-temporal convolution and an improved prototypical network for ECG classification
Jia Li, Dainan Gan, Xiaolin Zhan +6
PLOS One15 Oct 2026Peer-reviewedCC BY 4.0Full textCardiovascular medicinePLOS
Multi-task deep learning framework for brain tumor classification and segmentation using ensemble feature extraction
Retinderdeep Singh, Chander Prabha, Shahab Abdulla
Scientific Reports18 Oct 2026Peer-reviewedCC BY-NC-ND 4.0Full textNeurologyCrossref
An Explainable Disease-Aware Intelligent Histopathological Representation Learning Framework for Generalizable Breast Cancer Classification
Faizan Ahmad, Irfanud Din, Khalil Khan
Diagnostics (Basel, Switzerland)1 Oct 2026Peer-reviewedCC BYFull textGeneral medicineEurope PMC

Cite

BibTeX RIS

BibTeX

@article{sooraj2026temporal,
  title = {{Temporal deep learning framework for multi condition fault classification of three-phase induction motors}},
  author = {R. Sooraj and Senthil Kumar Ramu and R. Sitharthan and Satyam Kumar and Richu Zayan Thayyil},
  journal = {Scientific Reports},
  year = {2026},
  month = jul,
  doi = {10.1038/s41598-026-62262-3},
  url = {https://doi.org/10.1038/s41598-026-62262-3},
  note = {CC BY},
}

RIS

TY  - JOUR
TI  - Temporal deep learning framework for multi condition fault classification of three-phase induction motors
AU  - R. Sooraj
AU  - Senthil Kumar Ramu
AU  - R. Sitharthan
AU  - Satyam Kumar
AU  - Richu Zayan Thayyil
PY  - 2026
DA  - 2026/07/01
JO  - Scientific Reports
DO  - 10.1038/s41598-026-62262-3
UR  - https://doi.org/10.1038/s41598-026-62262-3
AB  - Induction Motors (IMs) are a very important group of electrical appliances used in industry, due to their strength, versatility, high reliability, and cost-effectiveness. However, due to various faults, there can be many issues, such as production downtime, energy losses, and increased maintenance costs. This paper employs Deep learning (DL) frameworks, such as Simple Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and Gated Recurrent Unit (GRU), for the detection and categorization of IM. The novelty of the proposed work lies in the unified multi-sensor, multi-class, and multi-model DL framework with comprehensive cross-architecture evaluation for robust IM fault diagnosis under eight different operating scenarios, including startup transients and phase-removal faults, load-based faulty and healthy conditions. The results show that the DL is effective in electrical fault identification, using the classification accuracies of RNN—85.14%, LSTM −90.19%, BiLSTM—92.24%, and GRU—94.19%. The proposed approach can be used to differentiate between healthy and faulty operating conditions, which aim to highlight the effectiveness of DL for intelligent condition monitoring in induction motor drive-based machines.
SN  - 2045-2322
ER  -

Abstract

Related

Cite