Open Rotor Phase Fault Detection in Wound-Rotor Induction Machines Using Signal Texture and Shallow Neural Networks - Université de Picardie Jules Verne Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Open Rotor Phase Fault Detection in Wound-Rotor Induction Machines Using Signal Texture and Shallow Neural Networks

Résumé

Studies related to fault detection in induction motors have taken a next step as machine learning techniques are becoming popular as the industries adapt to Industry 4.0 and make provisions for Industry 5.0. In relation to that, this paper proposes a texture-based feature estimation coupled with a shallow neural network for detection of open rotor phase in wound rotor induction machines using only 3-phase current signals. After signal conditioning of the acquired experimental data and calculation of contrast definitions (texture-based), optimized shallow multi-layer perceptron neural networks have emerged to be the best classification model with respect to its other neural variants. The model selection has been done based on overall architecture, classification accuracy, confidence in probability predictions, time complexity and least number of trainable parameters.
Fichier non déposé

Dates et versions

hal-04520281 , version 1 (25-03-2024)

Identifiants

Citer

Rahul Kumar, Giansalvo Cirrincione, Mehdi Taherzadeh, Humberto Hénao. Open Rotor Phase Fault Detection in Wound-Rotor Induction Machines Using Signal Texture and Shallow Neural Networks. 2023 IEEE International Conference on Advanced Systems and Emergent Technologies (IC_ASET), Apr 2023, Hammamet, Tunisia. pp.1-6, ⟨10.1109/IC_ASET58101.2023.10151214⟩. ⟨hal-04520281⟩
15 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More